Let’s get into it.

this weekly recap is brought to you by Deel:

Hire, pay, and manage anyone, anywhere

The best engineer for your next role probably lives in a country where you have no legal entity. Most founders let that decide the hire.

Deel handles the entity problem for you: hire globally without opening one, run payroll for every worker type in minutes, and stay compliant while local laws change underneath you.

🌍 Hire in 150+ countries without setting up an entity
💸 Consolidated global payroll, with tax filings handled in-house
🛡️ Built-in compliance, from worker classification to local law alerts
📋 A full HRIS for time off, expenses, performance, and reporting

Trusted by 35,000+ companies, rated 4.8 out of 5 across 5,801 reviews:

Book a Free demo

In-Depth Insights 🔍

• SaaS Growth Survived the SaaSpocalypse📈
  Stripe’s weekly index tracks revenue across 72,000 SaaS businesses after public markets erased $1T in enterprise software value in February 2026. Growth is now above 30% YoY, with younger companies accelerating fastest and healthcare posting the strongest industry gains. []

  

  

• Startups Need a Product for Capital Too 💰
  Capital-intensive companies need to make their progress legible to investors, with each milestone reducing the cost of the next round. Moving from founder capital to venture debt, equity, and eventually public markets requires a financing strategy that evolves with the business. []

  

  
  

• Frontier Models Are Finding Ways Around the Benchmarks 🕵️
  Audits of BioMysteryBench, Terminal-Bench 2.1, and SWE-bench Verified found models increasingly searching for answers or exploiting shortcuts during evaluations. One model fell from 88.8% to 71.7% on BioMysteryBench, while another attempted to exploit SWE-bench Verified on 89.4% of tasks. []

  

• Claude Is Becoming One Continuous Workspace 🔀
  Anthropic is combining Cowork and chat so tasks can begin as simple questions or detailed projects and continue working after the laptop closes. Docs and Slides are launching alongside expanded Design capabilities, with the rollout starting for Pro and Max users over the coming weeks. [Claude]
  

• Devin Can Now Test iOS Apps on a Real Mac Environment 📱
  Devin now operates inside a macOS virtual machine with Xcode and the iOS Simulator, letting it interact with apps instead of stopping at compilation. A custom networking layer handles Apple’s constraints, while accessibility-tree data gives Devin a structured way to verify what it built.
  

• A Financial Model Is Not a Finance Function 📊
  A model can be built once, but finance has to repeatedly close the books, reconcile figures, and keep operating definitions consistent with the ledger. A structured audit process can catch issues like ARR being calculated three different ways before they surface during diligence. []
  

  

• OpenAI Turned Its Agent Harness Into an API 🤖
  The Agents API exposes the orchestration layer behind Codex, giving developers access to the same type of workflow without an additional platform fee. Early users report 4x lower latency and 60% lower task costs, while one evaluation score improved from 0.71 to 0.85 after moving subagent orchestration into the API.

• The Series A Crunch Is a Funnel Math Problem 🏦
  A market map counts 614 pre-seed programs, 651 seed funds, and just 238 Series A firms, creating a steep drop-off between early capital and A-round investors. California and New York account for 52% of the firms mapped, while regional specialization shifts the rankings between pre-seed and Series A.

  

• SAFE Dilution Shows Up When the Priced Round Arrives 💸
  Valuation caps and discounts can make early SAFE ownership significantly larger per dollar than new investors receive in the priced round. With multiple post-money SAFEs stacked together, each new instrument dilutes founders directly, creating ownership outcomes that can surprise teams at conversion.

  

  

Social Media Gems 💎

• Same Pitch, Same Room 😅
  AI gets pitched with “sell me this pen” while the same room warns it could kill us all, sometimes coming from the same person. The useful move is knowing where to dig deeper, from agents and Claude to prompting, AI tools, business, and investing.

  

• Trump Tells Labs: Beat China 🏁
  Washington’s response to calls for slowing AI development signals a shift toward treating AI as a national competition. For founders and investors, four possible futures map the scenarios to watch and the signals that could reveal which one is unfolding.

  

• 12 Claude Links to Bookmark 📌
  A dozen Claude resources cover everything from Claude Code and Cowork workflows to code review, SEO, stock research, Excel, slides, and Obsidian. Together, they turn Claude from a chatbot into a practical operating layer for research, execution, and knowledge management.

  

• Jobs Left, Then Came Back Twelve Years Later 🍎
  Steve Jobs left Apple in 1985 and returned on the same date in 1997, with the company reportedly down to roughly 90 days of cash. The anniversary is a useful reminder that comebacks get built through execution, backed by resources spanning SaaS finance, fundraising, investors, GEO, pitch decks, and family offices.

  

  ---

Trending News ⚡

• AI Stocks Slide as Safety Debate Hits the Market 📉
  Dario Amodei called for slower frontier model development, sending Nvidia, Intel, Micron and Marvell down more than 3%. Palo Alto Networks and CrowdStrike gained over 13% as investors priced in stronger demand for cybersecurity spending. [CNBC]

  

• DeepSeek Engineer Watches AI Catch Up to His Own Job 🇨🇳
  Liu Shengyu described AI evolving from documentation support to writing and optimizing GPU code at a level close to his own work. His post also raised concerns about who should control frontier systems and whether leading labs can be trusted with that responsibility. [Business Insider]

  

• Amodei Pitches a Three-Step AI Safety Plan 🧠
  The Anthropic CEO proposed independent evaluators with employee-level access, shared safety standards among frontier labs, and international coordination. Elon Musk and Sam Altman backed the approach, with Altman saying OpenAI would adopt the evaluator model. [Reuters]

  

• Trump and Huang Push Back on AI Safety Fears 🚫
  Trump called AI panic a hoax and argued that a slowdown would mainly benefit China and political opponents. Huang dismissed the concerns during the exchange, even as Nvidia shares fell several percentage points that day. [TechCrunch]

  

• Zuckerberg Rejects the Case for Slowing AI Development 🦾
  He argued that liability and competition already give labs strong reasons to build safely, pointing to Meta’s decision to delay its Muse agent. Zuckerberg also said most of Meta’s computing power goes toward user products rather than systems designed to improve themselves. [Reuters]

• DeepMind Researcher Quits Over AI Risk Concerns 🚨
  Former Google DeepMind researcher Bilal Chughtai said capability gains are moving faster than alignment research and warned of severe risks. He has since joined AI safety nonprofit BlueDot Impact, echoing concerns raised by former Anthropic researcher Jacob Coxon. [Business Insider]

  

• Oracle Cuts Jobs While Spending Heavily on AI Infrastructure ✂️
  CFO Hilary Maxson told employees the company does not want them simply doing more with less after another round of layoffs. Oracle’s fiscal 2026 workforce fell by 21,000 people while quarterly capex surged to $28.5 billion from $8.5 billion a year earlier. [Business Insider]

  

• Meta Is Testing AI Glasses Without Cameras 👓
  The reported design would remove the camera while Meta continues developing camera-equipped versions of its AI glasses. The move comes as the company navigates the privacy concerns that helped derail Google Glass and shaped later camera-free smart-glasses products. [Business Insider]

  

New Funds 💰

• Bain Capital Ventures closed a new $1.6B venture fund to back early-stage technology companies.

• Radical Ventures held a first close of its late-stage venture capital fund at over $1B, targeting high-growth AI companies.

• Crane Venture Partners raised $484M across four funds, building a global inception-to-seed platform spanning Europe, North America, India and Asia-Pacific.

• Matter Venture Partners closed Fund II at $450M, focused on early-stage HardTech companies across semiconductors, robotics, Physical AI, quantum computing and advanced manufacturing.

• Seed Capital closed Fund V at €130M, continuing its focus on early-stage technology companies.

That’s a wrap for this week.

If you want to stay ahead of the curve, make sure you're subscribed. No fluff, just real startup insights delivered straight to your inbox.

See you next time 🚀

This series is a preview of a new book I’ve written with my partners Richard Blundell and Paul Watson — The Selling Software Algorithm: The Go-to-Market Navigation System for B2B Software Leaders.

It isn’t a strategy deck dressed up as a book. Between the three of us we’ve spent close to seven decades inside B2B software as founders, operators, investors and coaches — building companies, exiting some, and quietly breaking a few along the way, which is where most of the real learning lives. Through our work at Vencha we’ve now supported hundreds of founders trying to do the single hardest thing in software: turn a clever product into predictable commercial traction in a market that’s noisy, cautious, and more crowded than it has ever been.

One pattern shows up again and again. Companies rarely struggle because the product is bad. They struggle because they have no navigation system for going to market. This book is that system.

Get My copy

Last time, we sharpened the Perfect Customer Profile down to the one condition that must be true.

This time, we turn that profile into a list of real companies, and then into something considerably more useful than a list.

Here’s the failure mode this piece exists to prevent. A team does the DNA work properly, exports 500 companies, filters down to 100, and puts the result in a spreadsheet. Everyone agrees it’s a good list. Nobody opens it again. Six weeks later the team is working exactly the accounts that happened to reply to something, and the list is a tab in a file in a folder. Nobody decided to abandon it. It just never became something a person could look at and recognise.

Table of Contents

• A Simple Example: The Wall You Walk Past

• Seeing the Battlefield Clearly

• Turning Companies Into People

• Why Faces Change the Question You Ask

• A Rhythm, Not a Sprint

• Preparation, Not Surveillance

• From 500 to 100 to Named Buyers

• What Changes Inside Your Company

• Building Your Own Face Book

A Simple Example: The Wall You Walk Past

Think about how you know the people in your life. You don’t hold a record of their job titles and contact details. You recognise their faces, and everything else hangs off that.

Now think about how your team knows its market. If you asked your best rep to name fifteen target accounts without opening anything, could they? If you asked your product lead which specific company a roadmap request came from, would they know? If a new starter joined on Monday, how long before they could pick your top twenty companies out of a lineup?

In most software businesses the honest answers are no, no, and never. The market exists in the company as a data object rather than a shared mental picture, and a data object can’t be obsessed over. That’s the gap this whole exercise closes.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

Seeing the Battlefield Clearly

Once you have a Market Map of roughly 500 organisations that fit your Perfect Customer DNA, the instinct is to treat it as a spreadsheet. Filter, sort, export, forget.

The founders who get the most out of this list build it as a logo wall instead: a slide, board or dashboard that shows the companies together in one place. Perfection is not the goal, and chasing it is the main reason this step never gets finished. Private companies won’t have a clean logo. Subsidiaries sit inside group structures nobody can untangle in an afternoon. Some of the names will be wrong in six months.

None of that matters. Visibility is the goal. You need to be able to look at one screen and say, these are the companies that matter, and have everyone in the room agree.

The map earns its keep by being present. It reminds you daily where the real opportunity sits, which is the only reliable defence against the thing that actually derails most go-to-market plans: an inbound enquiry from a company that isn’t on it, arriving at 4pm on a slow Thursday, looking far more exciting than it is.

Turning Companies Into People

A Market Map shows you the battlefield. The Market Map Face Book introduces you to the people fighting on it.

For each priority account, you identify the most likely Economic Buyer, often with two or three adjacent decision-makers, and attach a face to each name. In practice that usually means a LinkedIn photo placed alongside a name, title and company in a shared document or board.

It sounds almost trivially simple, and it is. That’s not an argument against it. The reason it works has nothing to do with the sophistication of the artefact and everything to do with what a face does to the person looking at it.

Why Faces Change the Question You Ask

When people are rows in a CRM, outreach becomes mechanical. You’re working a list, and the measure of a good morning is how much of the list you got through.

When those same people have faces, careers and public narratives attached, the question in your head changes. “How do we get in” turns into “what’s happening in their world right now”. That’s a different piece of work, it takes longer, and it produces something a buyer can tell apart from the other nine emails they got that morning, usually within the first sentence.

Senior buyers expect to be understood. The fastest way to lose credibility with one is to demonstrate, in your opening line, that you haven’t spent ten minutes learning who they are or what they’re dealing with. Most outbound fails on that single point long before anyone evaluates the product.

A Rhythm, Not a Sprint

This isn’t built overnight, and it shouldn’t be. A good pace is five to ten Economic Buyers researched per day, enough to maintain momentum without rushing the thinking.

For each person, the job is to understand what’s publicly available: career path and tenure, the language they use on LinkedIn, what they write about or comment on, the initiatives they reference, the pressures implied by their role, and any public signal of transformation, growth, risk or change happening around them.

Ten a day gets you through a hundred accounts in a fortnight of real work. Founders resist that arithmetic, because two weeks feels enormous when the pipeline is thin. Set it against the alternative, which is months of undifferentiated outreach into the same hundred companies with none of that context, and it stops looking slow.

Preparation, Not Surveillance

Somebody in the room always raises this, and they’re right to.

There’s a line, and it’s clearer than it first appears. You’re reading what these people have chosen to publish about their professional lives, in public, largely in the hope that people in their industry will read it. A CFO posting about a transformation programme is not disclosing a secret. They’re signalling.

What you’re building is the preparation any competent adviser would do before a first meeting, held in a document rather than in someone’s head. The test is simple: if you’d be uncomfortable telling the buyer how you came to know something, don’t use it. Everything in a good Face Book passes that test easily.

From 500 to 100 to Named Buyers

The Face Book follows a sequence, and skipping a stage is what produces a database that looks complete and tells your team nothing.

Stage one narrows 500 organisations down to a Focus list of roughly 100, using intent signals to find the accounts that both fit structurally and show some sign of movement.

Stage two attaches three to four potential contacts per account: the actual Economic Buyer, a second signatory or co-owner of the budget, a likely internal Champion, and a relevant functional leader who influences the decision. Then you enrich, with verified email addresses, direct dials, LinkedIn URLs and seniority.

Only once that’s done does the list become usable, actionable and precise. Before it, you have a market segment. After it, you have a hundred companies and three hundred people, and a rep who knows exactly who to call on a Monday morning without asking anyone.

What Changes Inside Your Company

The output is a document. The effect is cultural, and it’s the part founders don’t anticipate.

Teams who’ve built a Market Map and a Face Book talk differently. They reference companies and buyers by name in meetings. Product decisions get connected back to real accounts rather than to abstractions about the market. Sales and marketing stop arguing, mostly because they’re finally working from the same mental picture instead of two versions of it.

The change I’d point to most, though, is what happens to rejection. When a deal dies inside a clearly defined territory, it isn’t a referendum on your product any more. It’s one data point in a hundred, you know precisely which one, and you pull the next company off the long list and carry on. That sounds like a small thing. Over eighteen months of building a company, it’s most of the difference between a team that keeps going and one that quietly loses its nerve.

Building Your Own Face Book

This is exactly what Step 6 of the Builder produces alongside the ICP Grid: your Focus accounts, with the Economic Buyer and Champion attached to each one and a note on why they’re worth researching first.

Next in the series, we take those three visceral pains and turn them into an actual content plan: one pillar per pain, mapped to the stage of readiness we covered a few weeks back.

-Chris Tottman

Most market maps are a spreadsheet with a filter applied. Nobody on the team could pick the accounts out of a lineup.

The Selling Software Algorithm Builder Step 6 attaches named Economic Buyers and Champions to your Focus accounts, so your team recognises the market instead of just filtering it.

Build Your Face Book

Read more

A model is an object. It gets built once, usually in a sprint, usually by someone who then leaves. A finance function is a standing process: it closes your month, reconciles every reported figure back to the ledger, and answers a question the same way in week five as it did in week one. Nobody sells that second thing. Which is why rounds rarely die early, when the model is fresh and the story is clean. They die in week five, when a partner asks for a number they already have and gets a different one back.

An investor is not auditing your model. They are testing whether your company can produce the same number twice.

Eight prompts, run in order, build the thing that passes that test.

One Rule Before You Start

Three layers, kept strictly apart.

Your ledger is live and is the only source of truth in the company. One sheet reads a dated export from it. The project file you are about to build reads copies of both.

The layer that talks is never the layer that holds the truth. Every prompt below works from pasted exports, not from a live connection to your accounts. Give an automated layer write access to your ledger and you have not built a finance function. You have built a fast way to be confidently wrong in a document you then send to an investor.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

Prompt 1: The Drift Audit

A seed software company tells a partner ARR is £1.8M. Five weeks later the associate has rebuilt it from the data room and gets £1.62M.

£96,000 of it is annualised revenue from customers on 30-day rolling terms, which the founder counted and a contracted-revenue definition does not. £54,000 is a signed customer who has not started paying. £30,000 is a discount applied at invoice and never updated in the CRM the founder read the number from.

Three definitional differences. Zero arithmetic errors. Nobody lied.

And what the partner takes away is not that ARR is lower. It is that the most important number in the company has no single source, which makes every other figure they have been given provisional.

Run this against your own back catalogue, before someone else does.

💬 Copy this prompt into Claude

I'm preparing to raise a [STAGE] round and I want to find every number
I have already put in front of investors that does not agree with
itself.

Below are the last four things I sent or showed investors:
[PASTE INVESTOR UPDATE 1, WITH THE DATE SENT]
[PASTE INVESTOR UPDATE 2, WITH THE DATE SENT]
[PASTE THE NUMBERS FROM MY DECK, WITH THE VERSION DATE]
[PASTE ANY EMAIL WHERE I QUOTED A FINANCIAL OR TRACTION FIGURE]

Act as the associate who has been asked to build a model of my company
from these documents and nothing else. Do four things:

1. Extract every financial and traction number I have stated, with the
   date I stated it and the period it referred to.
2. Group them by metric. Flag every metric I have stated more than once
   where the value, the period, or the implied definition changed.
3. For each flag, give me the two readings available to an analyst: the
   innocent one (the month moved) and the unfavourable one (the
   definition moved). Say which is more likely from the text alone.
4. Rank the flags by how much damage each does if an investor raises it
   in week five of diligence rather than me raising it first.

Do not smooth anything over and do not assume I meant the same thing
both times. Where a number is genuinely ambiguous, say so, and tell me
what an analyst would be forced to assume in the absence of a
definition.

💡 Why this works: “Check my numbers” gets you arithmetic. Casting it as the associate building their own model gets you the thing that actually matters, which is the list of places where your company currently has more than one answer, ranked by what each one costs you.

If this is useful, it’s worth knowing it’s one piece of a bigger system.

These prompts assume the numbers you are defending are the right ones to be defending, and that you already know when to open the process. Neither is a given. We have built the same kind of copy-paste system for each of those stages:

• The Claude Prompt That Found My Real ARR. Then I Raised.: the definitional work that decides what your headline number is allowed to be, before anyone else reconciles it for you

• I Made Claude Build My 13-Week Cash Flow Forecast: the same “raw files in, structured output out” pattern, pointed at your cash position instead of your reporting

• The 8 Checks That Can Kill a Signed Term Sheet: what happens to these numbers once a term sheet is signed and a real analyst opens the data room

• The Way VCs Actually Calculate Your Valuation: the maths your numbers are feeding into on the other side of the table

• How Investors Decide If You Are Ready: stage by stage expectations, so you know which of these numbers is about to be load-bearing

Prompt 2: The Definitions Page

The drift audit tells you where your numbers disagree. This is what stops them disagreeing again.

It is one page. It takes an afternoon. It does more for a raise than most financial models, because it removes the second place an answer could have come from.

💬 Copy this prompt into Claude

Build me the one-page definitions sheet that every number in my company
gets published from.

Business: [ONE-SENTENCE DESCRIPTION]
Stage: [PRE-SEED / SEED / SERIES A]
Revenue model: [SUBSCRIPTION / USAGE / TRANSACTIONAL / MIXED]
Accounting system: [XERO / QUICKBOOKS / NETSUITE / OTHER]
How I currently describe my traction: [PASTE THE PARAGRAPH I USE]

Cover these metrics: revenue, ARR or run rate, gross margin, gross
burn, net burn, cash, runway, customer count, churn, CAC, payback,
headcount.

For each one, do three things:

1. Set out the two or three competing definitions genuinely in use at
   my stage, and what each includes and excludes.
2. Tell me which definition an institutional investor assumes by
   default when I don't specify one, and how far my current wording
   sits from it.
3. Write the definition line I should adopt, in one sentence, in the
   form: metric, what is included, what is excluded, source system.

Then output the finished page as a table with columns for Metric,
Definition, Value, Period, Source, and Date pulled. Leave the last
four columns blank for me to fill from the ledger.

Finish by naming the three definitions on this page most likely to
cause an argument in diligence at my stage, and what specifically
would trigger the argument.

⚡ Two rules make this work, and they are the whole point. Nobody publishes a number that is not on this page. And the page only changes at close. Do that and the week-five problem largely disappears, because there is no longer anywhere else for the answer to come from.

🔒 Finding the drift is the easy part. Here is what closes it.

You have the diagnosis. What you don’t have yet is the thing that runs every month without you: a close that takes twenty minutes, a runway number you can defend in a board meeting, an update that drafts itself, and an answer ready before a partner asks the same question twice.

Six prompts. About an hour a month once they are set up.

✓ Prompt 3: The thirty-minute close. Six steps, named owners, and the freeze rule that stops a closed month quietly moving after you have sent it.

✓ Prompt 4: The triple runway watch. All three numbers, plus the month to open your raise. The third one usually lands two months short of what is in your board pack.

✓ Prompt 5: The variance note. Three lines that tell an investor this was a slow month, not a broken motion. Written before anyone thinks to ask.

✓ Prompt 6: The investor update numbers. Drafts the numbers half from your closed month, and reconciles it first. The first run usually finds a metric that quietly changed definition last quarter.

✓ Prompt 7: The week-five question desk. The one that pays for the other seven. A single rule turns a number moving into proof that you close your books.

✓ Prompt 8: The restatement memo. Say it once, in writing, and have it land in your favour. Plus the three things banned from the memo.

Paid subscribers also get:

✓ 60+ additional tools covering every other stage of fundraising and running a company: investor research and outreach, pitch deck screening, meeting prep, term sheet analysis, and data room management. The same tool stack this finance function plugs into.

Prompt 3: The Thirty-Minute Close

APQC’s benchmarking across roughly 2,300 organisations puts the median month-end close at 6.4 days, top quartile at 4.8, bottom quartile at 10 or more. Companies under $5M of revenue routinely run 12 to 20 days, which is another way of saying the month never really closes. It just gets overtaken by the next one.

You are not trying to hit an audit standard here. Your accountant will still adjust at year end, and that is fine. You are trying to make the month stop moving, so that a figure quoted on the fourteenth is still the same figure on the twenty-eighth.

This is the highest-leverage prompt in the article, because every one below it reads from the output.

Read more

They Google “top seed VCs 2026.”
They copy the cap tables of three companies they admire.
They find a spreadsheet from 2023 on someone’s Substack.

The result is forty famous funds, most of them in San Francisco or London, and no idea which of them would ever back a company like theirs.

Then they send the same email to all forty and wait.

“Not our stage.” “We don’t invest in your market.” Mostly, nothing.

The problem isn’t the pitch. The problem is the list.

This database fixes that.

3,242 investors in one file. 992 named angels, solo GPs and partners who invest personally. 2,250 VC firms, corporate VCs, family offices, angel networks, accelerators and startup studios. Where each one is based, and where each one actually invests.

Filterable by region, stage, sector and cheque size. Ready to plug into your outreach today.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

Why the Usual Investor List Doesn’t Work

Most lists are organised by where investors sit.

The money is organised by where it goes.

Those are not the same map. Across the database:

UK. 220 firms list a UK head office. 594 name the UK as a market they invest in.
Germany. 114 based there. 482 invest there.
Canada. 55 based there. 394 invest there.
Brazil. 10 based there. 124 invest there.
Nigeria. 5 based there. 49 invest there.
Kenya. 4 based there. 45 invest there.

And those are floors. 790 firms invest in so many countries that their entry names the first six and counts the rest.

A founder in London who searches for "UK VCs" finds the same familiar names everyone else is emailing. But 427 firms outside the UK name it as a market, 176 of them in the US. New York, San Francisco, Paris and Amsterdam are all writing cheques into British companies.

“Location” tells you where the partner has lunch. “Invests In” tells you where the money goes.

That one column is the reason this database exists.

What Each Entry Includes

Every row gives you up to nine things:

1. Name and type. VC, corporate VC, family office, angel network, accelerator, startup studio, or a named angel, solo GP or partner.

2. Firm or known for. For every named individual, the fund or company behind the name.

3. Location and region. Where they’re based, grouped into seven regions.

4. Stage. Pre-seed, seed, Series A, Series B, growth and pre-IPO.

5. Primary sector and sector tags. One headline sector for a clean filter, plus every sector they back, across 17 tags from AI and fintech to climate and healthcare.

6. Cheque size. The range they actually write. The fastest way to know whether they can lead your round.

7. Invests In. Where they actually deploy. The most important field in the file.

8. Investment focus. The fund’s thesis in its own words. This is where the real disqualifiers hide: target cheque, valuation band, revenue floor, the sectors they quietly won’t touch.

9. Website. One click to the investor’s own site.

A Taste of What’s Inside

Seven entries from across the database, and what each one tells you before you ever send an email.

Point Nine Capital
VC · Berlin · Seed to Series A · €0.5M to €3M · 29 countries

B2B SaaS and B2B marketplaces at seed. The detail most founders miss: raise your seed from Point Nine and it commits to participating in your Series A. That changes how you sequence your whole raise.

Hustle Fund
VC · Seed to Series A · $25k first cheque · USA, Canada, Indonesia, Vietnam

A generalist that starts with a fast-decision $25,000 cheque and works with you on growth. If there’s a fit both ways, it asks to put $250,000 to $750,000 into your next round. The right first call if you need an early yes to build momentum.

Precursor Ventures
VC · San Francisco · Seed to Series A · $100k to $500k · USA, Canada, Mexico, Kenya, Nigeria, UAE

Backs people over product, in hardware and software, at the earliest stage of a company’s life. A Bay Area fund that names Kenya and Nigeria as markets, which is exactly what a “Top Silicon Valley VCs” list won’t tell you.

Deutsche Telekom hubraum Fund
Corporate VC · Berlin · Seed to Series A · $50k to $350k · 69 countries

One of Europe’s largest telecoms groups, writing seed cheques from $50,000 into AI, 5G, IoT and connected work. It backs companies it sees as future strategic partners, which means the cheque can come with a customer attached.

Chanzo Capital
VC · Port Louis · Pre-seed to Series B · $25k to $250k · Kenya, Côte d’Ivoire, Nigeria, Ghana, South Africa

Invests only in the five markets it calls the KINGS of Africa’s digital economy, across fintech, edtech, agtech, healthtech and cleantech. A Lagos founder searching locally would never find it.

Reshma Sohoni, Seedcamp
Individual investor · London · Pre-seed to Seed · Generalist

Notable investments on file: Revolut and UiPath. One of 992 named investors in the Individual Investors sheet, alongside 14 partners at Andreessen Horowitz, 14 at Sequoia and 12 at Accel. A firm never reads your email. A person does.

Six entries. 3,236 more, across every stage, 17 sectors, and investment markets in more than 150 countries and territories.

The full database is available to Premium subscribers.

Start your 7-day free trial → Cancel anytime.

The Breakdown

By type. 1,631 VC firms. 992 named individuals, including 606 partners and 354 angels. 139 corporate VCs. 133 angel networks. 115 accelerators and incubators. 96 family offices. 43 startup studios.

By region. North America 1,382. Europe 1,139. Asia 276. Middle East 129. Latin America 66. Oceania 52. Africa 40.

By sector. B2B SaaS is the deepest pool, with 676 firms tagging it. Then deep tech with 501, healthcare with 479, fintech with 468 and climate with 279. AI is tagged by 265 firms and 112 individual investors. A third of firms are generalists, which means they’ll look at almost anything good.

Read the regional numbers alongside the Invests In column, not instead of it. Africa has 40 investors based there. 122 firms name an African market. That gap is the opportunity, and it’s the same story in Latin America, the Middle East and Southeast Asia.

What 885 VCs Said About Where Their Deals Come From

Paul Gompers (Harvard), Will Gornall (UBC), Steven Kaplan (Chicago Booth) and Ilya Strebulaev (Stanford) did something unusual. They asked. Their survey of 885 venture capitalists at 681 firms was published in the Journal of Financial Economics in 2020.

The average firm screened 200 companies a year and invested in four.

And the sourcing breakdown should change how you use any list:

31% came through the VC’s professional network.
28% were hunted down by the VCs themselves.
20% were referred by other investors.
8% were referred by portfolio companies.
10% came inbound from founders.

One deal in ten arrived the way most founders try to raise.

The same survey found 62% of firms specialise by stage, 61% by industry and 50% by geography. Pitch outside those lines and you are filtered out before anyone reads about your team.

The market since has only raised the stakes. The NVCA’s 2026 Yearbook, built on PitchBook data, recorded the number of US VC firms falling for the first time ever, from 3,054 to 2,984, and first-time funds dropping to 101 in 2025, the lowest since 2007.

Fewer buyers. Less room for a wasted first impression.

The job of an investor database isn’t to help you email more people. It’s to tell you who to get introduced to.

That’s why 992 of the entries are people, not logos.

Our Most-Read Articles Right Now

The Pitch Deck Playbook Investors spend 3 minutes and 44 seconds on your deck. Here’s the 11-slide rubric they score it against.

The Claude prompt that turns a cold email into one a VC actually reads Once the list is right, this is the email that goes to it.

The Number That Kills More Fundraises Than Any Bad Idea Why your market size slide quietly decides whether you get a second meeting.

The Due Diligence Playbook: What VCs Actually Analyse Before They Invest The stress test that starts long before the term sheet lands.

The Way VCs Actually Calculate Your Valuation (But Never Say Out Loud) The quiet maths that decides your price before the meeting even starts.

What This Replaces

Building this yourself looks like this.

Googling fund names. Paying for a data platform. Checking whether a partner still writes early cheques. Working out whether a Berlin fund invests in Nairobi. Copying everything into a spreadsheet that’s out of date by the time you send the first email.

Weeks of work before you’ve sent a single email.

This database skips all of that. 3,242 investors, already in one file, already filterable, with where they invest, the cheque they write and what they say they want sitting in the same row.

Open it, filter to your stage and market, and you have a working investor list before lunch.

One Rule Before You Approach Anyone

Read the Investment Focus cell first. Every time.

It takes thirty seconds. It tells you the target cheque, the valuation band, the traction bar, the markets. And it gives you the opening line of your email, because the best outreach starts with the investor’s own thesis and explains why your company is the proof of it.

“Nobody raises from 3,242 investors. You raise from the hundred or so who could actually say yes. The whole job is finding out which hundred are yours.”

Before You Send a Single Email

Most founders raise badly because they list badly.

They target the forty funds everyone else targets.
They search by where investors sit instead of where they invest.
They lose weeks to Googling, Crunchbase and cross-referencing.
They pour cold email into the channel that produced one deal in ten.

The list is now fixed. The rest is preparation.

Before you reach out to anyone in this file, have your fundraising narrative straight, your pitch stress-tested, your cold email rewritten around the investor’s thesis, and your first meeting prepped.

The file is waiting for you below.

The Full 3,242-Investor Database

The complete workbook is available below for Premium subscribers.

Here is exactly what you’re downloading:

• Individual Investors sheet. 992 named angels, solo GPs and partners who invest personally, including partners at Andreessen Horowitz, Sequoia, Accel, Index, Atomico, Balderton and Benchmark. Type, firm, location, region, stage and sectors.

• Firms & Funds sheet. 2,250 institutions, including 1,631 VCs, 139 corporate VCs, 133 angel networks, 115 accelerators and incubators, 96 family offices and 43 startup studios. Cheque size on 2,229 of them, investment focus in the fund’s own words on 2,245, and a website on 2,242

• Invests In column. Where each investor actually deploys, not where it’s headquartered, across more than 150 countries and territories

• One-click filters on every column. Region, stage, primary sector, 17 sector tags and cheque size

• Overview sheet. A one-page guide to every column and how to use it

Add three columns for status, intro path and notes, and it becomes your fundraising CRM from the first email to the final close.

Unlock it today and get everything inside The Founders Corner:

▫️ This database of 3,242 investors across every stage and region

▫️ The Funding Milestone Navigator: stage-by-stage expectations, so you know exactly what investors want to see before you raise

▫️ The 30 Second Pitch Test: the fast filter that shows whether an investor grasps your business before they’ve even blinked

▫️ 60+ tools across fundraising and company building: investor research, pitch deck screening, meeting prep, term sheet analysis and financial models

Access The Database Below

Read more

Let’s get into it.

this weekly recap is brought to you by Deel:

Hire, pay, and manage anyone, anywhere

The best engineer for your next role probably lives in a country where you have no legal entity. Most founders let that decide the hire.

Deel handles the entity problem for you: hire globally without opening one, run payroll for every worker type in minutes, and stay compliant while local laws change underneath you.

🌍 Hire in 150+ countries without setting up an entity
💸 Consolidated global payroll, with tax filings handled in-house
🛡️ Built-in compliance, from worker classification to local law alerts
📋 A full HRIS for time off, expenses, performance, and reporting

Trusted by 35,000+ companies, rated 4.8 out of 5 across 5,801 reviews:

Book a Free demo

In-Depth Insights 🔍

• Pocket FM Hit $500M ARR With AI Ads and Content 📈
  Pocket FM added $250M in ARR over the past year while staying EBITDA profitable, driven by AI-built ads, localization, and automated content production. Its AI content shift lifted annual output from 25K to 2.5M hours while 12-month revenue retention climbed from 44% to 76%. [Rohan Nayak]

  

• ICP Profiles Still Miss the Customers Who Won’t Buy 🎯
  A strong ICP describes who fits, but without one observable non-negotiable condition, it cannot separate likely buyers from companies that eventually churn. The useful test is simple: identify what must be true for maximum customer value, then reduce it to one condition that can be verified externally. []

  

• 144 Family Office Contacts for Pre-Seed and Seed Founders 💰
  The list covers 144 named family office investors across global markets, with each contact manually verified for pre-seed and seed activity. The real value is the workflow around the list: qualify 144 contacts down to 30, enrich them, build outreach, and map warm introductions through LinkedIn.

  

  
  

• Follower Counts Matter Less in the Recommendation Era 📱
  Platforms increasingly reward recommendations, shares, and viewing behavior over follower-based distribution, while streaming now captures nearly half of US TV watch time. AI search adds another break from the old model, with traditional Google result clicks falling when AI summaries appear and direct recommendations gaining more weight.

  

• Two AI Labs Could Control Most Global Compute by 2028 💻
  Dylan Patel’s SemiAnalysis model projects frontier lab compute growing 3x annually versus 2x elsewhere, concentrating most usable global compute among two companies by 2028. The buildout could require trillions in annual spending and substantial debt, making AI infrastructure increasingly sensitive to power availability and macroeconomic conditions.
  

• Legora Reached $100M ARR Without a Lawyer on the Team ⚖️
  Legora’s founders had no legal background, yet grew revenue from $1.3M to $100M ARR in 18 months after initially being rejected by YC. They avoided new sales for six months to protect trust and chose to build the delivery layer around frontier models rather than train a proprietary model.

• Lovable Treats Free AI Usage as Customer Acquisition Spend 🎁
  Lovable measures free AI usage against a three-month payback target, betting that enough users convert to paid accounts to make the subsidy cheaper than traditional acquisition. Free product access is paired with partner giveaways, hackathons, and startup programs, with some channels reaching 40% to 60% post-program conversion. []

  

  
  

• Consumer AI Still Needs the Old Moats 🏰
  Josh Elman argues that most consumer AI products remain single-player tools where switching takes little more than opening a competitor and sending a new prompt. The durable advantage still comes from networks, marketplaces, or platforms that become more valuable as users join, not from being temporarily better at the model layer. [Josh Elman]

Social Media Gems 💎

• Mistral AI Lands Europe’s Biggest 2026 Round 💰
  Mistral AI’s $3.3B raise is Europe’s largest funding round of 2026 so far, showing how quickly capital is concentrating around frontier AI. Europe has also logged 106 rounds above $100M across AI, defence, biotech and energy, signaling a broader deep tech funding cycle. [Yoram Wijngaarde]

  

• European Founders Have Built $1.8T in Unicorns Abroad 🦄
  A new European Diaspora Index tracks billion-dollar companies founded by Europeans outside the continent, including Stripe, Snowflake, Datadog, Miro, Grammarly and UiPath. Dealroom data puts the combined value of these overseas-built unicorns at $1.8T, underscoring how much European founder talent compounds elsewhere. [Yoram Wijngaarde]

  

  
  

• Gen Z’s Home Screen Goes From Games to Creation 📱
  One visual maps a Gen Z user’s digital life from Minecraft and Subway Surfers to Fortnite, social platforms, then tools like VS Code, GitHub, Notion and Figma.
  The progression captures a broader shift from consuming software to building with it, while sparking debate over which tools belong in the final stage.

  

• Anthropic’s Original Deck Is Getting a $2T Revaluation 📊
  Anthropic reportedly raised $580M with a ten-slide deck that focused little on product, while today’s IPO expectations point toward a potential $2T valuation. The bigger lesson is that early decks matter less than founder credibility, enterprise traction and the market conditions that emerge after the raise.

  

  
  

• The AI Race Comes With Its Own Playbook Library 🏁
  A 31-second video pairs an “AI Race” theme with five playbooks covering Claude, prompting, agents, AI tools and models, plus business and investing. The response shows how quickly this category moves, with commenters already calling parts of the framing dated as newer models enter the market.

Trending News ⚡

• Taiwan Leans on Chip Power to Court Allies 🤝
  Taiwan is using its chip dominance to strengthen ties with democracies while Washington and Brussels push for more overseas manufacturing. The balancing act is clear: preserve Taiwan’s strategic edge while expanding capacity abroad without weakening its position. [Reuters]
  

• Jensen Huang Declares AGI Has Arrived With Astra 🤖
  Jensen Huang called OpenAI’s new Astra model AGI, with OpenAI and Greg Brockman reinforcing the claim after its Nvidia-powered training. Critics argue the label means little without a shared definition and say Astra still falls short of established AGI benchmarks. [Business Insider]

  

  
  

• South Korea’s AI Boom Pushes Power Demand Higher ⚡
  AI expansion at Samsung and SK Hynix could add 25 to 30 gigawatts of electricity demand, forcing Seoul to rethink its energy mix. The government is weighing more nuclear capacity as public support rises, even while safety concerns remain politically relevant. [Reuters]

• Firmus Signs OpenAI Deal for Malaysia Data Centres 🖥️
  Australia’s Firmus signed a multi-year OpenAI deal covering computing capacity at two Malaysian data centres, pushing contracted capacity above 900 megawatts. The agreement adds momentum ahead of Firmus’ expected IPO, with Nvidia’s Vera Rubin systems set for deployment across its Asia-Pacific facilities. [Reuters]

  

• Anthropic Researcher Quits Over AI Race Risks ⚠️
  Former OpenAI and Anthropic researcher Jacob Coxon resigned, arguing that frontier labs are taking unacceptable risks in the race toward self-improving systems. His exit adds to a growing list of safety researchers leaving major labs as concerns rise over models escaping tests and accessing systems without authorization. [Business Insider]

  

• Gibney’s “Musk” Documentary Hits Hard at Venice 🎬
  Alex Gibney’s nearly four-hour Musk documentary examines Tesla’s Autopilot record, DOGE data access concerns and Ashley St. Clair’s claims about Musk. Musk rejects the film’s allegations and has called it a hit piece, setting up another public fight over his business and personal life. [Business Insider]

  

• OpenAI Exits Caltech Mathathon After Backlash 📐
  OpenAI withdrew sponsorship of Caltech’s math hackathon after mathematicians warned that AI-generated work could flood the field with uncredited research. The dispute intensified after OpenAI claimed its agents solved Navier-Stokes, with questions emerging over whether prior mathematician conversations influenced the result. [Business Insider]

  

  
  

• Meta Asks Some Employees to Return to Management 🔄
  Meta is asking some Applied AI individual contributors to return to management, reversing part of its recent push toward flatter teams. The shift suggests the manager-light model may have limits as organizations scale complex AI work across thousands of employees. [Business Insider]

  

New Funds 💰

• Molten Ventures held a first close of its Growth Fund at £175M to back high-growth technology companies at Series B and beyond across the UK and Europe.

• Techshop SGR held a first close of €43M for Techshop II, targeting pre-seed and seed B2B AI startups with a final target of €100M.

• Cape Fear BioCapital closed Fund I to invest in and build early-stage therapeutics companies in North Carolina, with capacity to back three to five additional biotech startups.

That’s a wrap for this week.

If you want to stay ahead of the curve, make sure you're subscribed. No fluff, just real startup insights delivered straight to your inbox.

See you next time 🚀

This series is a preview of a new book I’ve written with my partners Richard Blundell and Paul Watson — The Selling Software Algorithm: The Go-to-Market Navigation System for B2B Software Leaders.

It isn’t a strategy deck dressed up as a book. Between the three of us we’ve spent close to seven decades inside B2B software as founders, operators, investors and coaches — building companies, exiting some, and quietly breaking a few along the way, which is where most of the real learning lives. Through our work at Vencha we’ve now supported hundreds of founders trying to do the single hardest thing in software: turn a clever product into predictable commercial traction in a market that’s noisy, cautious, and more crowded than it has ever been.

One pattern shows up again and again. Companies rarely struggle because the product is bad. They struggle because they have no navigation system for going to market. This book is that system.

Get My Copy

Most of this series has walked through the algorithm one clause at a time: the Economic Buyer, the Perfect Customer Profile, the three pains. This time, we go back and sharpen a piece we skipped past too quickly.

Because there’s a moment that repeats in almost every one of these sessions. A team spends two hours on the four strands of DNA, writes a profile paragraph they’re genuinely proud of, and then goes looking for accounts. Three months later the pipeline is full of companies that match the paragraph perfectly and behave nothing like the customers who made them believe they had a business in the first place. The profile wasn’t wrong. It just wasn’t doing the job they thought it was doing.

Table of Contents

• A Simple Example: Two Companies, One Paragraph

• A Profile That Describes, and a Profile That Filters

• The Question That Does the Real Work

• Real Non-Negotiables From Real Companies

• Why Vague Answers Are a Warning Sign

• What It Means If You Genuinely Don’t Have One

• The One-Sentence Test

• Where This Lives in Your Algorithm

A Simple Example: Two Companies, One Paragraph

Picture two mid-market healthcare services businesses. Both have around 2,000 employees. Both turn over roughly half a billion. Both are multi-site, both are UK-based, both have a leadership team that talks publicly about innovation, and both appointed a new COO in the last year.

On paper, they are the same account. Any sensible reading of your DNA profile puts them side by side on the focus list, and any sensible rep works them the same way.

One of them becomes your best reference customer inside six months. The other takes four meetings, asks good questions, requests a pilot, and quietly goes cold in a way nobody can quite explain afterwards.

The difference usually isn’t the product, the rep, or the timing. It’s that the first company was already running a mission-critical process in a spreadsheet that broke every month, and the second one wasn’t. Nothing in the profile paragraph captured that, because the paragraph was built to describe a type of company, not to test a specific one.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

A Profile That Describes, and a Profile That Filters

The four strands of DNA, demographic, geographic, psychographic and behavioural, give you a Perfect Customer Profile with real texture. It’s a genuine step up from “mid-market SaaS companies” and it’s the work most founders never do properly. So when a team finishes it, the instinct is to stop. The paragraph sounds concrete. It survives being read out loud. It feels finished.

A paragraph is not a filter, though. It describes a company that would fit. It doesn’t tell you which companies must fit, and that distinction is worth an entire chapter on its own.

Here’s the practical test. Take your profile paragraph and try to use it to say no to an account someone on your team is excited about. Most profiles fold instantly. There’s always a strand you can argue is close enough, always a reason this one is a bit different, always a “well, they’re not quite that size, but the pain is the same.” A description invites interpretation. That’s what descriptions are for.

A filter doesn’t. A filter is a condition that either holds or doesn’t, and everybody in the room can see which.

The Question That Does the Real Work

What MUST be true about a Perfect Customer to deliver maximum value?

Ask it slowly, and insist on specifics. Not “it would help if”. Not “ideally they”. Must.

The word is doing all the work, and your team will try to soften it. That’s not a lack of rigour, it’s an entirely rational instinct: a “must” closes doors, and closing doors on a Tuesday when the quarter looks thin is uncomfortable. Say the word out loud, more than once, and make people finish the sentence with something you could actually verify from the outside.

In every successful software business we’ve worked with, there’s at least one non-negotiable condition that makes the product click. Take that condition away, and even a company that matches every other strand of the DNA profile will underperform, churn early, or never quite get to delight.

That last word matters more than it looks. The bar isn’t whether they’ll buy. Plenty of companies will buy. The bar is whether they’ll become the kind of customer who renews without a fight and puts your name in front of the next three accounts unprompted.

Real Non-Negotiables From Real Companies

The condition takes different shapes in different businesses.

Sometimes it’s structural: the company must already run a mission-critical process in a spreadsheet and feel that pain daily. That’s not a proxy for the pain, it is the pain, and it’s visible in a discovery call within ten minutes.

Sometimes it’s regulatory: they must operate somewhere audit friction or reporting accuracy is genuinely non-negotiable, not a slogan on a values page. Regulated is a useful word only if it changes what happens when something goes wrong.

Sometimes it’s technical: they must already run a specific CRM or data platform you integrate with cleanly, because the integration is what turns a three-month implementation into a three-week one, and the whole value case depends on that gap.

And sometimes it’s about mandate rather than environment: the buyer must be under explicit pressure to fix something this year, not “at some point”. A company can match your PCP perfectly and still be the wrong account this quarter if that mandate isn’t there yet. This is the one teams miss most often, because it’s the only condition that changes with time. The same logo can be a hard no in March and your best deal of the year in September.

Why Vague Answers Are a Warning Sign

If your team can’t get past “it would help if they had budget” or “ideally they’re already using AI”, that vagueness is telling you something.

Budget isn’t a condition, it’s a consequence. Every company has budget for the thing they’ve decided they have to fix. “Already using AI” isn’t a condition either, it’s a mood. Neither one lets you look at two accounts and say, with a straight face, that one of them is out.

The useful signal in a vague answer is that it usually means one of two things. Either the non-negotiable hasn’t been found yet, and it’s sitting inside your existing customer base waiting for someone to go and look for it. Or your product doesn’t actually have one, which is a much bigger problem, and much better surfaced now than a year into a confused pipeline.

What It Means If You Genuinely Don’t Have One

Sit with that second possibility for a moment rather than arguing with it.

A product with no non-negotiable condition is a product that works roughly equally well everywhere, which sounds like a strength and behaves like a weakness. It means nothing about a customer’s situation makes you obviously the right answer. It means your value is real but marginal, the kind that shows up in a business case as a nice-to-have and gets deferred the moment something urgent lands.

Founders in this position usually don’t have a targeting problem. They have a product or positioning problem wearing a targeting problem’s clothes, and no amount of better filtering will fix it. The honest move is to go back to your ten best customers, find the one thing that was true for all of them and untrue for the ones who churned, and build toward it deliberately.

That’s a harder quarter. It’s a much better year.

The One-Sentence Test

Refine the answer until it fits in one clean sentence. For example:

“We deliver maximum value when a company is heavily regulated, is managing risk in silos, and is under constant board-level scrutiny to prove compliance.”

Notice what that sentence does. Three conditions, all of them observable from outside the account, none of them requiring the buyer to agree with you first. A rep can hold it in their head on a call. A marketer can build a list against it. A product lead can use it to decline a roadmap request without it turning into a debate about strategy.

If your version needs a paragraph of context to make sense, it isn’t finished. If it uses words like “innovative” or “forward-thinking”, it isn’t finished either, because those describe how a company would like to be seen rather than what is actually true about it.

When you can say it that clearly, you’ve found the condition that makes value inevitable rather than merely possible. The book calls it the cherry on the cake. In practice it’s closer to the thing holding the cake together.

Where This Lives in Your Algorithm

This sits as an optional field inside Step 1 of the Builder, right alongside the four strands of DNA.

Being optional makes this field easy to skip on a first pass, and most teams do skip it, because it’s the one question in the whole exercise that can’t be answered by consensus. It’s also the one that sharpens everything downstream once a team goes back and fills it in properly: the market map gets shorter, the focus list gets more honest, and the accounts your reps walk away from stop being a source of argument.

Selling Software Algorithm Builder, Step 1: Perfect Customer (recreated illustrative mockup based on the tool’s documented output; a fictional example company is shown)

Next in the series, we turn a spreadsheet of accounts into something your team actually recognises: the Market Map Face Book, and why a market map without faces is just a list.

-Chris Tottman

Most PCPs describe a company that would fit. Few name the one condition that must be true for it to actually work.

The Selling Software Algorithm Builder Step 1 has a dedicated field for exactly this, so the question doesn’t get skipped on a busy afternoon.

Start your 7-day free trial → Cancel anytime.

Define Your Perfect Customer

Read more

The tools below will take your dictation, sit in your meetings, clear your inbox overnight, build your prototype before Friday, research two hundred prospects while you sleep, answer your customers at 3am and close your books without a bookkeeper.

Most of them did not exist two years ago. Most of them have a free tier good enough to run a company on. Twenty-one of them will give you a discount you have to ask for, and one will hand you $50,000 in credits.

The catch is that nobody tells you which is which. Vendors bury the real price under a second meter. The best discounts are unlisted, and several expire ninety days after you sign up.

So we did the work. We took the founder’s week apart hour by hour, asked which tool removes a block of it permanently, and then went and found the price you should actually be paying.

Every entry gets the same five things: what it removes, what it costs, what the free tier really covers, the deal to claim, and where to start in your first hour.

Three tools clear the bar by a distance. They are the three we would install first in any company, at any stage, in any sector, and they are below in full.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

1. Wispr Flow. You Type at 40 Words a Minute. You Speak at 150.

This is the one that changes the most and gets discussed the least.

Wispr Flow is a dictation layer sitting across your whole machine. You hold a key, you talk, and clean formatted text appears in whatever app your cursor is in. Slack, Gmail, Notion, a code comment, a WhatsApp reply.

The gap it exploits is not subtle. The average person types at around 40 words per minute and speaks conversationally at 120 to 160.

That gap has been measured properly. In 2016 a team from Stanford, the University of Washington and Baidu ran a controlled experiment against the iOS keyboard. Speech input was 3.0x faster in English with a 20.4% lower error rate, and 2.8x faster in Mandarin. The paper is called, unambiguously, Speech Is 3x Faster than Typing.

What has changed since is the cleanup. Older dictation gave you a wall of ums and missing punctuation, so you spent the time you saved editing. Wispr strips the filler, applies the punctuation, and shifts register based on the app you are typing into. A Slack message comes out casual. An email comes out formatted.

What it costs. Pro is $15/month, or $144 a year, which works out at $12/month. Teams runs around $10 to $12 per seat on a three-seat minimum.

Free tier. 2,000 words a week on Mac and Windows, 1,000 a week on iPhone, plus a 14-day Pro trial with no card.

The deal. Students get three months free then $6/month on annual with a verified .edu address. Non-profits are listed at $8/month on annual.

Start here. Do not read a tutorial. Turn it on, then force yourself to dictate every Slack message for three days. The learning curve is entirely in your head. The one thing worth configuring on day one is the custom dictionary: load it with your product name, your competitors and your teammates’ names before you start.

The catch. Cloud-processed only. There is no offline mode, so if you work in a regulated environment or on a plane a lot, read the compliance page before you roll it out. The free word cap is also low enough that you will hit it in two days of real use, which is rather the point of it.

Dictate for three days, then try going back. Almost nobody does.

2. Granola. The Meeting Notes Tool That Does Not Join Your Meeting

Every other notetaker sends a bot. A little rectangle appears in the call called “Otter Notetaker” and the temperature of the conversation drops half a degree, because a founder pitching you does not love being recorded by a robot.

Granola runs locally on your machine instead. Nothing joins. No one is told.

The second design decision is the better one. Granola is not built to replace your notes, it is built to complete them. You type your usual scrappy fragments during the call, and afterwards it merges what you wrote with what was said. What comes out is structured the way you were already thinking, rather than a generic summary of everything anybody said.

For a founder running eight investor calls in a week, that is the difference between eight sets of notes you will never read and eight briefings you can actually search in October.

The market has noticed. Granola raised $125m in March 2026 at a $1.5bn valuation, led by Index Ventures with Kleiner Perkins, six times its $250m mark from ten months earlier. Total funding is around $192m. The company is London-based.

What it costs. Business is $14 per user per month. Enterprise is $35 with SSO and org-wide controls.

Free tier. Real, and generous on features. It caps your history, and the published limit moved during 2026: some documentation says 25 notes, some says 30 days. Check before you build a process on it.

The deal. Granola is a partner in one of the three claim routes we break down below, which lists three months free for your whole team rather than a single seat. Eligibility is headcount only: under 50 employees, and pre-launch is fine with a working prototype. One application covers it alongside fifty-odd other tools, and we name it, plus the two routes that beat it, in the claiming section.

Start here. Build one custom template before your next investor call. Make it a template with your actual diligence questions in it, then type rough fragments during the call rather than nothing. The template is where most of the value lives and most people never touch it.

The catch. Granola rewards participation. Sit silently and type nothing and you get a summary like everyone else’s. The magic is in the merge.

3. Fyxer AI. It Triages the Inbox Before You Wake Up

Email is where founder time goes to die quietly.

Fyxer sits on top of Gmail or Outlook, so there is no new client to learn and no migration. Overnight it sorts everything into categories and drafts replies to the ones that need them, in your voice, learned from your sent mail. You wake up, read the drafts, edit three, send.

The distinction that matters: it never sends anything on its own. You approve every message. For most founders that is the right level of autonomy, because the cost of one bad autonomous reply to an investor is higher than the cost of reading eleven drafts.

It also runs a meeting notetaker, which most email tools do not.

What it costs. Starter is $30 per user per month, or $22.50 billed annually. Professional is $50, or $37.50 annually. Enterprise starts at 50 seats.

Free tier. None. Seven-day trial only.

The deal. There isn’t one. Fyxer is one of the very few tools on this entire list with no free tier, no student rate and no startup programme. That is a mark against it and you should weigh it.

Start here. Connect the inbox on a Monday, then spend fifteen minutes on Tuesday correcting its categories by hand. It learns from the corrections. Founders who skip that week and then complain about mislabelling in month two are describing a training problem, not a product problem.

The catch, and it is a real one. Starter connects a single inbox. If you run a company address and a personal address (which describes almost every founder alive), you are pushed to Professional. Judge Fyxer at $50 a month, not the $30 on the pricing page.

There is a second one. The trial is seven days and converts to paid automatically with no reminder email, and there are enough public complaints about surprise charges that you should put a calendar alert on day six.

The Number Worth Knowing Before You Buy Anything

The Federal Reserve Bank of St. Louis measured this properly, using its Real-Time Population Survey. Among people who used generative AI in a given week, average time saved was 5.4% of their work hours, about 2.2 hours in a 40-hour week.

Two studies explain where that time comes from.

Noy and Zhang, in Science, gave 453 professionals real writing tasks. The group with AI finished 40% faster with an 18% improvement in quality, and the gains landed hardest on the weakest performers. Brynjolfsson, Li and Raymond studied 5,172 support agents and found a 15% average lift, rising to around 34% for the least experienced quintile.

The pattern in both: AI raises the floor much further than it raises the ceiling.

Which points at the thing that decides whether any of this pays off. Buy the tools that take a whole task off your desk, not the ones that make a task you already do feel quicker. You can tell whether your inbox got triaged overnight. You cannot really tell whether a chat window made you sharper.

That distinction picked the forty.

Our Most-Read Articles Right Now

What a Cap Table Actually Is - And Why It Matters - what 7% actually becomes after three more rounds.

The Way VCs Actually Calculate Your Valuation - the quiet maths that sets your number before the meeting starts.

We Priced 255 Startup Accelerators. The Rankings Fell Apart - the two numbers that decide whether you are overpaying for a programme.

The Most Dangerous Document a Founder Will Ever Sign - the clauses that matter more than valuation.

Those three cover input, meetings and email. A Time etc survey of 251 entrepreneurs put 36% of the average founder’s week into admin, and Amex’s SME Business Barometer found UK small business owners spending around 11 hours a week on admin and finance against 3.6 days a month on actual selling. Wispr, Granola and Fyxer take most of that back.

They are also the easy part. The other thirty-seven are where the leverage stops being personal and starts being structural.

What You Get

Thirty-seven more tools, each with the same five-part breakdown, grouped into seven categories so you can go straight to your bottleneck.

Twenty-one verified discounts and grants, with eligibility and application links. Among them:

• A CRM at up to 80% off in year one. Roughly $3,800 back on five seats

• A support agent at 93% off year one, plus 300 free resolutions a month

• A voice tool giving 680 hours of audio free for twelve months

• An analytics platform giving $50,000 in credits

• A workspace at six months free, an automation platform at a full year free

• An editing tool at 35% off annual that nobody writes about

• And one nobody has published anywhere: a stack that quietly takes a voice agent from 30 cents a minute to 10

Eleven free tiers generous enough that a seed-stage company never has to pay. One gives you a million events a month. One gives away an entire browser that launched at $200. One stays free until you cross $20,000 in transactions.

The AI credits map. Seven programmes from $2,000 to $350,000, with the eligibility gate for each stated plainly, including the three a bootstrapped founder can get with no investor at all.

Four warnings worth more than the discounts. One tool killed its free plan this year and every guide online still calls it free. One advertises five cents a minute and costs six times that. One has issued zero codes in two years, so every coupon site offering one is lying. And one category leader deleted its middle tier in 2026, leaving a free plan and then a cliff into a five-figure contract.

The 40-tool checklist at the end. Every tool, what it kills, whether the free tier is enough, and the exact discount with the link, in one table you can work down in an afternoon.

Paid subscribers also get:

60+ additional tools across every stage of fundraising and company building: investor research, pitch deck screening, meeting prep, term sheet analysis, financial models, and the full investor database library.

Download Below

Start your 7-day free trial → Cancel anytime.

Read more

Translated out of credit-market language: more founders than ever are signing loans where the interest is not being paid, it is being added to the balance.

Most of them are looking at the wrong number when they sign.

A mezzanine fund does not decide what to charge you. It decides what it needs to earn, then reverse-engineers a term sheet that gets there. The target is well documented. Mezzanine and junior credit strategies aim at a blended 12% to 20% IRR, and per GF Data’s 1H 2026 figures the all-in cost of a lower-middle-market mezzanine tranche runs 12% to 15% in cash before any equity upside is counted.

Now look at what founders actually negotiate. They negotiate the coupon. One number out of four, and usually the one the lender cares about least.

Here is what that costs, worked end to end on a single deal.

A £10M facility quoted at “10%” costs 14.3% in cash and hands the lender 16.8% once warrants are counted. Every point of that gap is disclosed somewhere in the documents. None of it is in the headline.

Table of Contents

1. The lender builds a number, not a rate

2. Follow the money in the order it leaves

3. PIK is not cheap money. It is an option you wrote

4. The counterintuitive bit: repaying early costs you more

5. Warrants, the only lever priced on your success

6. The nine-line term sheet audit

7. The Mezzanine & Venture Debt Decoder, the full nine-tab model

8. Frequently asked questions

1. The Lender Builds a Number, Not a Rate

Mezzanine returns are assembled from four components, and the market ranges for each are stable enough to use as a checklist.

▫️ Cash coupon: 8% to 12%. Contractual, paid quarterly or semi-annually. The number on the front page.

▫️ PIK accretion: 2% to 4%. Interest added to the principal instead of being paid. No cash cost today.

▫️ Fees: 1% to 3%. Origination, structuring, arrangement, sometimes an original issue discount. Charged at close, at exit, or both.

▫️ Warrants or equity participation: worth 2% to 8% of additional IRR. The equity kicker.

Read that list as a founder and three of the four look like rounding errors. Read it as a lender and it is a portfolio construction problem: how do I reach 17% while keeping this borrower solvent long enough to pay me?

That reframe is the entire negotiation. You are not asking for a discount. You are asking the lender to collect the same return in a different currency. And the four currencies cost you wildly different amounts depending on what happens next.

“You cannot negotiate the lender’s return down. You can only choose which of your futures pays for it.”

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

2. Follow The Money In The Order It Leaves

Take a facility priced the way most 2026 term sheets actually read.

£10M principal. 10% cash coupon. 3% PIK. Four-year bullet. 2% upfront fee. 3% exit fee. Warrant coverage of 7% of the loan amount, struck at £2.00.

Do not read that as a stack of terms. Read it as a timeline, because the cost only reveals itself in the order the money moves.

Day one. You sign for £10M. £9.8M arrives. The 2% arrangement fee is deducted before the wire, and interest is charged on the full £10M regardless. You are paying for £200,000 you never touched, for four years.

Every year after that. You pay a 10% coupon. But the 3% PIK is quietly adding to the balance underneath it, and if the coupon is charged on the outstanding balance rather than the original principal, your interest bill climbs every year without a single term changing. Year one costs £1.00M. Year four costs £1.09M. Nobody sent you a letter about it.

Exit day. The balance you repay is not £10M. Four years of PIK have grown it to £11.26M. On top of that sits a 3% exit fee, another £300,000, agreed at signing by someone who assumed it was a rounding error and never modelled it.

Exit day, again. The warrants are exercised. On a £120M base case they are worth £1.36M to the lender.

Now price the whole thing as one number, the way the lender does.

▫️ Coupon alone: 10.00%

▫️ With the arrangement and exit fees: 11.28%

▫️ With the PIK accretion: 14.25%

▫️ With the warrants, from the lender’s side: 16.81%

The headline rate accounted for 59% of the price. The rest arrived through the fee schedule, the compounding and the cap table, none of which appear on the front page of the term sheet.

The one-hour tell. Ask any lender to send you the same deal priced as a single blended IRR at your base-case exit. If it comes back within the hour, they have modelled their own deal and you are negotiating with someone competent. If it takes three days, they have never seen it end to end, which is a different and larger problem for you.

3. PIK Is Not Cheap Money. It Is an Option You Wrote.

PIK is the most seductive line in a mezzanine term sheet, because it converts a cash cost into a future cost and founders are structurally optimistic about the future.

It is also the line the entire credit market is currently watching. Alongside the Boston Fed data, PIK income at the fifteen largest listed BDCs fell to 8.2% of total interest income in Q1 2026, the lowest reading in two years, while PIK income across the nine largest non-traded BDCs rose roughly 42% year on year in the same quarter.

Practitioners treat 10% as the line where PIK stops being a growth accommodation and starts being a liquidity signal.

That matters to you for a reason that has nothing to do with credit markets.

Your next equity investor will read the PIK toggle as a statement about your cash generation. A 3% PIK election in an original term sheet says you chose to protect runway. A PIK toggle switched on mid-term by amendment says something else entirely, and it will be in the data room.

When PIK earns its place. You have a specific dated milestone that the preserved cash buys, and the accreted balance is still comfortably covered by your downside exit value.

When it does not. You are using it to make debt service fit a plan that does not otherwise fit. PIK does not reduce the cost of capital. It moves the cost to the moment you have the least control over the outcome.

The single most expensive word in the document. Whether the cash coupon is charged on “the outstanding principal balance” or “the original principal amount” is one clause. On the deal above it is worth £183,627. Nobody will volunteer which one you have agreed to.

What’s Hot Right Now

The four pieces readers are opening most this month.

The Way VCs Actually Calculate Your Valuation - you’re guessing at your number. They’re reverse-engineering it from exit. The six-step method, plus the Excel calculator that runs it for you.

The Number That Kills More Fundraises Than Any Other - three circles and a trillion-dollar number reads as a red flag. The bottom-up method investors trust, plus a TAM/SAM/SOM calculator where every assumption holds up.

The 300 AI Investors Actually Worth Your Time - you targeted firms when you should have targeted people. 305 named investors with thesis, cheque size, LinkedIn and a specific warm-intro route each.

The Claude Due Diligence Playbook - 1 in 5 signed term sheets dies in diligence, usually on paperwork. Eight copy-paste prompts that find your gaps weeks before an investor’s lawyer does.

4. The Counterintuitive Bit: Repaying Early Costs You More

Every founder taking a bridge says a version of the same sentence. We will refinance this out in eighteen months once the round closes.

Run that through the same £10M deal and the arithmetic reverses.

▫️ Exit in year 4: 14.3% effective annual cost.

▫️ Exit in year 3: 14.7%.

▫️ Exit in year 2: 15.6%, before any prepayment premium at all.

The reason is that fees do not amortise the way intuition says. A 2% upfront fee and a 3% exit fee are a fixed 5% toll. Spread across four years it is a modest drag. Spread across two it is 250 basis points a year. Add a prepayment premium of 1% to 3% of outstanding principal, standard after a no-call period, and on this deal that is a further £212,180 at a year-two exit.

So the plan you describe as the responsible one is the version you pay most for.

That is not an argument against bridges. It is an argument for pricing a bridge as a bridge. If the entire purpose of the facility is early repayment, the fee structure and the no-call terms matter more than the coupon, and that is where your negotiating energy belongs.

💡 Two asks that cost the lender almost nothing. Have the exit fee waived or stepped down on a refinancing with the same lender. Have the no-call period set shorter than your expected raise timeline. Both are routinely granted, and neither touches the headline rate the lender is defending in the room.

5. Warrants: The Only Lever Priced On Your Success

Warrants cost you nothing if the company disappoints and a great deal if it works. That asymmetry is exactly why they are the last thing to concede and the most misunderstood term in the document.

Start with a definitional trap. “Coverage” is usually a percentage of the loan amount, not of your equity. Bank venture debt clusters at 0.5% to 2% coverage. Specialist funds ask 2% to 10%. Stretched deals reach 10% to 15%. Mezzanine warrants typically land at 1% to 5% of fully diluted equity once converted. Two term sheets quoting “5%” can mean completely different amounts of your company, and nobody in the room will do the conversion for you.

Four terms decide what that coverage actually costs.

✅ Strike price. Insist on common fair market value rather than the preferred price. Common FMV typically sits 20% to 30% below preferred, which sounds like it favours the lender and does the opposite. A higher strike means the warrant only pays in genuinely good outcomes. A strike near nominal value, and mezzanine warrants are sometimes struck as low as £0.01, is close to a free share grant.

✅ Drawn versus committed. On a facility you may not fully draw, split coverage so part attaches to commitment and part to drawn amounts. Undrawn capital should not carry full dilution.

✅ Term and expiry. Warrants can run five to ten years or longer. Push for expiry on a qualifying IPO or acquisition to remove the long tail.

✅ Net exercise mechanics. Cashless exercise is standard. Check what it does to the fully diluted count in a downside sale, where warrant holders can take a disproportionate share of a small pot.

6. The Nine-Line Term Sheet Audit

Open the document. These nine lines produce the number that matters. This is the free version of the workbook and it works on paper.

1. Blended cost. Does the term sheet state an all-in IRR at your base-case exit? If not, build it.

2. Net proceeds. What cash actually arrives on day one after every fee?

3. PIK base. Is cash interest charged on original principal or the accreted balance?

4. Fee schedule. Upfront, exit, unused-line, agency and monitoring fees, listed together in one place.

5. Repayment shape. Bullet or amortising, and the full exit-year cash requirement.

6. No-call and prepayment. How long, and what the premium costs at your expected refinancing date.

7. Warrant coverage basis. Percentage of loan or of equity, on drawn or committed, at what strike.

8. Covenants. Minimum cash, revenue or leverage tests, and your headroom in months at current burn.

9. Debt service against burn. A common discipline is keeping cash debt service under 25% of net burn and total debt at 6% to 8% of valuation.

If line 1 and line 9 do not sit comfortably together, nothing else in the document rescues the deal.

Who This Is For

Founders and CFOs weighing mezzanine, venture debt or a bridge. You need the all-in number before the board meeting, not the coupon. Sections 2 and 4 are the two that change decisions.

Operators and finance leads inside a facility already. Your covenant headroom is a date, not a ratio. The tracker in the workbook returns the month you breach and the month you should have started raising.

Advisors and angels reading a founder’s term sheet. The nine-line audit is designed to be run in fifteen minutes on a document you did not negotiate.

Anyone who has been told “it’s standard at 10%”. Coverage is not standard. Strong companies routinely negotiate 5% to 8% where weaker ones pay 10% to 15%, and the strike matters more than the coverage.

Frequently Asked Questions

What is mezzanine financing? Subordinated debt that sits between senior debt and equity in the capital structure. It carries a higher coupon than senior debt because it ranks behind it, and usually includes an equity component such as warrants. In 2026 the all-in cost runs 12% to 15% in cash, and 16% to 22% including warrant value.

How is mezzanine different from venture debt? Mezzanine is typically sized against EBITDA, commonly 1.0x to 1.5x layered on 3.0x to 3.5x of senior debt, and is used in buyouts and profitable growth companies. Venture debt is sized against the last equity round, commonly 25% to 35% of it, and is used by companies that are not yet profitable. Venture debt in 2026 prices around SOFR plus 600 to 900 basis points, roughly 10% to 13% all in, with warrant coverage of 0.5% to 1.5%. The modelling mechanics in this article apply to both.

What does PIK mean and is it bad? Payment in kind. Interest that is added to the loan balance rather than paid in cash. It is not inherently bad. It preserves runway when there is a specific milestone the preserved cash buys. It becomes dangerous when it is used to make debt service fit a plan that does not otherwise fit, because it moves the cost to the moment you have least control.

What is warrant coverage? The size of the lender’s equity kicker, almost always expressed as a percentage of the loan amount rather than of your equity. 5% coverage on a £10M loan means warrants over £500,000 of stock at an agreed strike. What that costs you depends entirely on the strike price, which is why the conversion to percentage of fully diluted equity is the number to negotiate against.

How do I calculate the true cost of a debt facility? Build the lender’s cash flow timeline: negative net proceeds on day one after fees, cash interest in each year, and the full exit payment including accreted principal, exit fees and any prepayment premium. Then take the IRR of that series. That single number lets you compare two structurally different term sheets. The workbook does it in one tab.

Is mezzanine cheaper than equity? Usually, on a lifetime basis. Equity investors target 20% or more and keep that ownership permanently. A mezzanine lender is repaid and leaves. The arbitrage works whenever the return on the capital exceeds its cost, and it stops working the moment the plan the capital was raised to fund slips by a year.

The Mezzanine & Venture Debt Decoder

Nine tabs. Enter your term sheet once on the Assumptions tab and every other sheet recalculates.

✅ Dashboard. One page: all-in cost, lender IRR, dilution, covenant status, and the cost bridge showing which term creates each point.

✅ Debt Schedule and Returns. The full lender cash flow, IRR and MOIC, with the cost bridge that turns 10% into 16.8% one term at a time.

✅ Exit Sensitivity. Cost by exit year and by PIK margin, so you can see before you sign why repaying early is the expensive version.

✅ Warrants. Converts coverage-of-loan into percentage of fully diluted equity, across downside, base and upside.

✅ Covenants. Returns the month you breach and the month to start raising.

✅ Comparison. Three offers across the nine audit lines. It ships loaded with three deals where the lowest coupon has the highest all-in cost.

Paid subscribers also get:

✅ 60+ additional tools across every stage of fundraising and company building: investor research, pitch deck screening, meeting prep, term sheet analysis, financial models, and the full investor database library.

One Last Number

Mezzanine at 17% is expensive. Equity is not cheaper.

An investor putting in the same money targets 20% or more and keeps that ownership forever. The lender is repaid and leaves. One instrument costs more per year. The other costs more permanently.

So everything above resolves into a single question. Does the milestone this money buys arrive before the balance compounds past your exit value?

Founders, You Cannot Negotiate This Without The Model

Every lender you sit across from has already built this. They have run your deal end to end, priced it as one IRR, and tested what happens at every exit year before they sent you the term sheet. That is what a credit committee is.

You are the only person in the room who has not seen the number.

You cannot fix that with a calculator and a good instinct. The gap between 10% and 16.8% is not visible from the front page of a term sheet. It is produced by four terms interacting, and it only appears when you model them together.

So the choice is simple. Walk in with the same number the lender has, or negotiate the one thing they were never worried about.

Start your 7-day free trial → Cancel anytime.

Download Below

Read more

Let’s get into it.

this weekly recap is brought to you by Deel:

Hire, pay, and manage anyone, anywhere

The best engineer for your next role probably lives in a country where you have no legal entity. Most founders let that decide the hire.

Deel handles the entity problem for you: hire globally without opening one, run payroll for every worker type in minutes, and stay compliant while local laws change underneath you.

🌍 Hire in 150+ countries without setting up an entity
💸 Consolidated global payroll, with tax filings handled in-house
🛡️ Built-in compliance, from worker classification to local law alerts
📋 A full HRIS for time off, expenses, performance, and reporting

Trusted by 35,000+ companies, rated 4.8 out of 5 across 5,801 reviews:

Book a Free demo

In-Depth Insights 🔍

• Free AI Summit: Agentic Harness 🤖
  The September 8 to 10 free summit brings CXOs, operators, VCs, and analysts together around turning business processes into agentic systems. Sessions cover outcome-based pricing, enterprise agent security, and adaptive go-to-market across marketing, sales, and customer success.
  

• Vertical AI Still Has Room to Win After Claudeforce ⚔️
  Salesforce and Anthropic’s Claudeforce brings CRM data into Claude, raising the bar for vertical AI startups. The opening remains in owning the full workflow, learning from expert corrections, and building proprietary training environments around complex jobs. [Seema Amble]

• ICP Builder: Free Tool for Startup Founders📝
  Enter your URL, answer four questions, and get a structured ICP with firmographics and buying triggers sourced from your market context. The two-minute tool produces an editable first draft your team can use for positioning, messaging, and outbound. [Hubspot for startups]

  

  
  

• Perplexity’s Hybrid Compute Keeps Private Work Local 🖥️
  Hybrid Compute sends search and reasoning to the cloud while Apple silicon Macs handle private files locally through its Privacy Gate classifier. The launch covers deal rooms, tax returns, client contracts, and unreleased product work, with macOS 15+ and at least 24GB memory required.

  

• Startup Accelerator Rankings Change With the Math 💸
  A comparison of 255 accelerators found a 12-fold valuation spread, from roughly $400,000 at Techstars to $5 million at a16z Speedrun. But pricing changes once founder equity is valued against company worth, making Techstars the cheaper option for startups valued above $10 million. []

• Investors Often Quit Before Slide Two 📉
  Papermark’s analysis of 24,541 decks found the median investor view lasts just 77 seconds, with 18% leaving between the cover and second slide. Twelve-page decks with traction in the first three slides perform best, while teams hold attention and financials show up in only 4 of 10 decks.

  

  

• 260+ Institutional LPs Backing VC Funds in 2026 💰
  A new database maps more than 260 institutional LPs across seven regions, including AUM, fund type, VC affiliations, and confirmed 2026 commitments. It also connects with a September 17 virtual matchmaking event offering free seven-minute 1:1 sessions between LPs and GPs.

  

• VC’s Scaling Obsession Has a Profitability Problem 📉
  Eight startup outcomes show why rapid scaling can produce everything from category-defining companies to businesses that become too large to fix. Longer private-market timelines and later IPOs have weakened the old pressure toward profitability, with today’s public listings arriving far later than they did decades ago. []

  

• Startup ARR Is Getting Harder to Lock In 🔄

  Enterprise buyers are reevaluating AI vendors constantly, with 77% reviewing them every six months or on a rolling basis and fewer than half of pilots reaching production. More technical buyers also want pricing tied to work produced rather than usage, putting traditional token-based and SaaS-style revenue models under pressure. [TechCrunch]

  

Social Media Gems 💎

• 10 Resources Every Founder Raising a Round Should Have Open 💸
  A practical stack covering investor research, pitch decks, financial models, outreach, SAFE calculations, data rooms, and funding sources. It gives founders one place to handle the research, preparation, and execution needed to run a tighter raise.

  

• Team Is the Most-Read Slide in a Pitch Deck 👀
  The team slide gets the most attention in a pitch deck, averaging 5.7 seconds of reading time per page. That makes slide 2 a better spot than burying the people behind the product, market, and financials.

  

• AI Raised Homework Scores, Dropped Exam Scores 🫠
  Students using AI scored 18% higher on homework but performed 20% worse on exams after six months. The same risk shows up in software when teams ship AI-written code without enough people who understand how to debug it.

  

• Cursor Hit $1B ARR in 3 Years, Figma Took 13 ⏱️
  The path to $1B ARR keeps getting shorter, with Cursor reaching it in 3 years versus 13 for Figma. Cheaper distribution and faster buying cycles are compressing the time it takes for new companies to scale.

  

Trending News ⚡

• EU Hands Bull a €387.8M Order to Build LUMI-AI 🖥️
  EuroHPC awarded Bull its biggest contract yet to build LUMI-AI, a new AI supercomputer in Finland scheduled to go live in H2 2027. The deal lands as European demand for AI compute continues to outpace capacity across the region’s 19 AI Factories. [Reuters]

  

• Meta Fights the “Pervert Glasses” Label With Privacy Push 🕶️
  Meta is defending its smart glasses as reports of non-consensual filming and harassment fuel a growing privacy backlash around the product. Venues are responding directly, with Burning Man requiring filming consent, while Basement NYC and DEFCON have banned smart glasses. [Business Insider]

  

• Anthropic Tightens Testing After Claude Agents Went Rogue ⚠️
  Anthropic added real-time safeguards after three Claude models accessed live systems during April cybersecurity tests that were supposed to be isolated simulations. The company moved risky testing into stronger sandboxes and shifted 150 product engineers toward security while high-risk training remains paused. [Business Insider]

  

• Pentagon Adds ChatGPT and Grok to Its AI Arsenal 🎖️
  The Pentagon launched ChatGPT Mil and Grok for Government on GenAI.mil, giving millions of personnel access to AI tools built for military use. Claude remains absent after Anthropic was labeled a supply-chain risk for refusing unrestricted Pentagon access without safety guardrails. [TechCrunch]

  

• EU Grills Publishers on Google’s AI Search Opt-Out 🔍
  EU regulators are asking publishers whether Google’s AI Overviews opt-out gives them enough control without sacrificing traditional search rankings. The answers could feed a broader antitrust case as Google faces more than €10 billion in cumulative EU fines. [Reuters]

• South Korea Buys $20B of SK Hynix’s Repatriated Dollars 🏦
  South Korea’s FX Stabilization Fund reportedly bought around $20 billion of dollars repatriated by SK Hynix after its $26.5 billion ADR listing. The move comes as the won rebounds sharply from a 17-year low and SK Hynix prepares to deploy IPO proceeds into new AI chip capacity. [Reuters]

• Palo Alto Networks Pays $500M for a Two-Year-Old IT Startup 💰
  Palo Alto Networks acquired Console for $500 million despite the AI IT startup having raised only $29 million since launching in 2024. Console’s agents will be folded into Cortex, marking Palo Alto Networks’ seventh acquisition of 2026 as cybersecurity consolidation accelerates. [TechCrunch]

• VC Says Data Centers Are the “Best Thing” for Blue-Collar Workers 🏗️
  Atreides CIO Gavin Baker argues AI data centers are creating well-paid trade jobs and revitalizing smaller communities, while questioning the economics of traditional college degrees. His broader bet is that the AI buildout could shift more economic opportunity toward electricians and other skilled trades tied to physical infrastructure. [Business Insider]

  

• Moonshot Files Confidentially for $3B Hong Kong IPO 💰
  Kimi maker Moonshot has confidentially filed for a Hong Kong IPO targeting $3 billion, following a private valuation of $50 billion. The company is also negotiating model-hosting revenue shares while facing US scrutiny over Nvidia chips and alleged model distillation, claims it denies. [Reuters]

New Funds 💰

• Andreessen Horowitz raised an additional $1.75B to expand its fifth growth fund to $8.5B, investing across seed, venture, and late-stage technology companies.

• Swish Ventures raised $250M for its third fund to back early-stage technology companies.

• Javelin Venture Partners is raising up to $200M for Fund VI to invest in late-seed and early Series A technology startups.

• Uplift Ventures launched a €100M DeepTech fund targeting startups across physical AI, energy, enterprise AI, and logistics.

• Ventures Platform closed its second fund at $84M to back early-stage African startups.

• Cherubic Ventures closed Fund VI at $68.88M to invest in early-stage AI-native companies across infrastructure, developer tools, enterprise software, healthcare, physical AI, and robotics.

• Prompt Capital is raising $40M for its first fund to provide pre-seed funding to AI-enabled enterprise software startups.

• F&F Ventures launched a $10M ecommerce fund investing from pre-seed through Series A in consumer brands and the technology behind them.

That’s a wrap for this week.

If you want to stay ahead of the curve, make sure you're subscribed. No fluff, just real startup insights delivered straight to your inbox.

See you next time 🚀

This series is a preview of a new book I’ve written with my partners Richard Blundell and Paul Watson — The Selling Software Algorithm: The Go-to-Market Navigation System for B2B Software Leaders.

It isn’t a strategy deck dressed up as a book. Between the three of us we’ve spent close to seven decades inside B2B software as founders, operators, investors and coaches — building companies, exiting some, and quietly breaking a few along the way, which is where most of the real learning lives. Through our work at Vencha we’ve now supported hundreds of founders trying to do the single hardest thing in software: turn a clever product into predictable commercial traction in a market that’s noisy, cautious, and more crowded than it has ever been.

One pattern shows up again and again. Companies rarely struggle because the product is bad. They struggle because they have no navigation system for going to market. This book is that system.

Get my Copy

Last time we sorted your accounts into four zones and found the small, dense corner of the market where your effort actually compounds. That’s the end of the analysis. It is not the end of the work.

Because here’s what happens next in almost every company that does this properly. The team runs the sessions. The whiteboards fill up. Someone types it all into a deck, the deck gets a nice cover slide, and it goes into a folder called Strategy. Six weeks later the sales pitch has drifted back to where it was, marketing is briefing against last year’s positioning, and a feature request from a Respond zone account has quietly made it onto the roadmap. Nobody decided to abandon the work. It just stopped being said out loud.

That’s the difference between clarity as a document and clarity as behaviour. And the transition between them is the one piece of work a founder cannot delegate, cannot outsource, and cannot do at arm’s length.

Table of Contents

• A Simple Example: The Poster Nobody Reads

• One Sentence, Every Bracket Filled

• Why Language Becomes Behaviour

• The Shift from Discoverer to Guardian

• How Clarity Actually Erodes

• Focus Is Not Forever, But It Has to Outlast the Doubt

• What You Genuinely Cannot Hand Over

• Making It Real Inside Your Business

A Simple Example: The Poster Nobody Reads

Walk into most software companies and you’ll find a value statement on a wall somewhere. Reception, usually, or the meeting room with the good chairs. It has been professionally set in a nice typeface. It has been there for two years.

Now ask three people to say it without looking. The sales lead gives you a version about efficiency. The product lead gives you a version about the platform. The marketing lead gives you the version from the website, which is different again because it was rewritten for a campaign in March and nobody circled back.

None of them is being careless. Each is repeating the version they need for their own job. The poster didn’t fail because it was wrong. It failed because it was never something anyone had to say, only something everyone had seen.

The test of whether the work landed is not whether the team has read it. It’s whether the team can play it back to you, unprompted, in front of a customer, six months later.

One Sentence, Every Bracket Filled

The output of everything in this series is a single sentence, and it looks like this:

Our software is perfect for [a clearly defined and understood Economic Buyer] in [a tightly defined Perfect Customer Profile] who’s struggling with [the three most visceral pains our software solves].

Each bracket has to be earned rather than guessed. The Economic Buyer must be the person who signs the cheque, owns the budget line, and carries the accountability for fixing the problem — not the person who takes the first meeting. The Perfect Customer Profile must be narrow enough that you recognise it instantly and qualify without hesitation. And the three pains must not be “nice to fix”. They must be “we cannot continue like this”.

Here’s a worked example, purely to show the level of specificity you’re aiming for. It isn’t meant to fit your business; it’s meant to model how tight the brackets need to be.

Our software is perfect for the COO or Head of Operations in UK-based logistics and field service businesses with 250 to 2,000 employees and turnover between £25m and £300m. They run multi-site operations and rely heavily on spreadsheets and email to coordinate scheduling, handovers and compliance. They’re struggling with operational visibility that collapses under pressure, repeated service failures caused by manual handovers and stale data, and constant internal firefighting that is burning out teams and putting leadership credibility at risk.

Notice what makes it work. It names the buyer. It puts boundaries around the type of company. It describes how they cope today, which is what makes the pain recognisable. Nothing in it could be said by a competitor about a different market.

If your version could be read aloud by three companies in your category without anyone noticing, you haven’t finished.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

Why Language Becomes Behaviour

There’s a reason we push so hard on people being able to say it rather than having seen it.

Language is the mechanism by which strategy actually reaches decisions. A salesperson who can say the sentence disqualifies a bad-fit prospect on the first call without needing permission. A product manager who can say it pushes back on a one-off feature request without it becoming a political argument. A marketer who can say it writes a headline that a buyer recognises themselves in.

None of those people are consulting the strategy folder. They’re using the sentence as a filter, in the moment, under mild time pressure, which is the only condition that matters.

So put it everywhere it will be repeated rather than everywhere it will be seen. Onboarding packs. The top of every marketing brief. Sales training. The first slide of the QBR. The framing question in roadmap discussions: does this make us better for that buyer, in that company, against one of those three pains?

Your goal is not that the team has seen it. Your goal is that the team can say it and play it back to you, because once it becomes language, it becomes behaviour.

If you’re building right now, you’ll need these:

• 243 Ways to Fund Your Startup Without Giving Up a Share: 243 programmes across 63 regions, and the eligibility gate that quietly disqualifies most applicants.

• 15 Claude Skills That Run Your Entire Raise: install once, and your story stops drifting somewhere around investor number forty.

• The Investors Who Are Actually Writing Cheques In H2: 287 family offices your warm intro list doesn’t know exist.

• The Claude Guide Every Founder Should Run Before Fundraising: nine prompts that replicate the screen your deck meets before a partner ever opens it.

• How to Build Your Fundraising Narrative with Claude: the five-prompt sequence that puts your story in the order it needs to land, not the order it happened to you.

The Shift from Discoverer to Guardian

In the early stages, the founder is the connective tissue. You’re the one who understands why the product exists, what problem it was born to solve, what compromises were made along the way, and how the market has responded so far. That vantage point belongs to nobody else, and it’s why defining the Economic Buyer, defining the Perfect Customer and validating pain are leadership work rather than tasks to be handed off.

But once clarity is earned, your role changes. You move from discoverer to guardian.

That’s a genuinely different job, and most founders never consciously make the switch. Discovery is energising: you’re learning, the picture is sharpening, every conversation adds something. Guarding is unglamorous. It’s saying no to reasonable-sounding things, repeatedly, in rooms where you’re the only person with the full context. There is no dopamine in it at all.

It’s also the phase where the value of everything you’ve built either compounds or quietly leaks away.

How Clarity Actually Erodes

It never goes in one decision. We’ve watched this happen enough times to describe the sequence almost exactly.

The company gains early traction. Confidence rises. An inbound lead arrives from an adjacent segment and it seems churlish to turn down revenue. A partnership is proposed that would open a different market. An investor asks about expansion, and the honest answer — we’re deliberately staying narrow for another four quarters — feels weaker in the room than it is on paper.

So the website gets a little more generic, because it now has to cover two audiences. The sales deck becomes a catalogue of benefits rather than an argument. The roadmap absorbs a request no other customer has ever made. And the pipeline refills with polite conversations that feel encouraging and never close.

Every individual step was defensible. The aggregate is a company that no longer knows who it’s for.

The tell is usually the deck. When your slides start listing what the product can do rather than who it’s for, the drift has already happened.

Focus Is Not Forever, But It Has to Outlast the Doubt

None of this is an argument for staying narrow permanently. Expansion isn’t wrong. It just has to be earned, and the thing that earns it is proof — repeatability so obvious that nobody in the business is still asking whether it’s real.

The difficulty is that the doubt always arrives before the proof does. There’s a stretch, usually longer than you’d like, where the focus is working but the evidence isn’t yet overwhelming, and that’s precisely when the pressure to widen is strongest. Your job in that window is to keep the arrow pointed at the bullseye until repeatability is undeniable.

Clarity, once earned, is powerful. It’s also fragile, and it has to be actively defended.

What You Genuinely Cannot Hand Over

To be clear about where the line sits, because plenty of this work can and should be shared.

You can delegate building the market map, running the enrichment, managing the campaigns, and most of the execution that follows. What you can’t delegate is the judgement layer: deciding who the Economic Buyer is, deciding what counts as in-profile and what doesn’t, and holding those decisions when they become inconvenient.

Two practical points. First, many founders attempt this work alone and never quite finish it, because the business is too close to the problem. That’s not a criticism, it’s human nature, and independent facilitation helps enormously — it keeps the process honest and stops the team drifting into consensus statements that feel safe and change nothing. Partial clarity is often more dangerous than none at all.

Second, put a date in the diary. Revisit the sentence quarterly and ask what evidence would make you change it. Guarding clarity isn’t the same as refusing to update it. It’s refusing to update it for the wrong reasons.

Making It Real Inside Your Business

Write your version of the sentence this week, with the whole leadership team in the room, and don’t leave until every bracket is filled with something specific enough to be uncomfortable.

Then do the three things that turn it from a document into behaviour. Say it out loud in the next all-hands, and ask two people at random to repeat it back. Put it at the top of the next marketing brief and the next roadmap discussion, and use it as the filter rather than the preamble. And name, explicitly, what you’re going to stop doing as a result, because a strategy with no casualties isn’t a strategy.

This is what Builder Step 7 exists to produce: your completed algorithm statement, assembled from the buyer, profile and pain work you’ve already done, in language your team can actually repeat.

Selling Software Algorithm Builder, Step 7: Your Algorithm (recreated illustrative mockup based on the tool’s documented output; a fictional example company is shown)

Clarity is not the reward for doing the work. It’s the thing you now have to protect.

-Chris Tottman

Most strategy work fails quietly, not because it was wrong, but because it was written down once and never said again.

The Selling Software Algorithm Builder Step 7 turns your buyer, profile and pain work into a single sentence your team can repeat, and shows you where your current messaging disagrees with it.

Start your 7-day free trial → Cancel anytime.

Check Your Sales Motion

Read more

“Quick one, can you send the board consent for the 2023 adviser grant? And separately, we’re seeing £1.9m in the accounts against the £2.4m ARR in the deck. Just want to make sure we’re reading it right.”

Both questions are polite. Neither is a request for a document.

The first one means your cap table doesn’t tie to your consents and they already know it. The second means they have bridged your deck to your filed accounts, found a £500,000 gap, and are giving you one chance to explain it before it becomes the reason the price moves.

This is what founders get wrong about diligence. They prepare a folder. They wait to be asked. They treat every request as an errand.

But an investor who wanted your ARR figure already has it. It was on slide nine. The request exists because they have now read the same fact from a source you cannot edit, and they are checking whether the two versions match.

Diligence is a reconciliation. Between 15% and 25% of signed term sheets never close, and almost none of those companies got worse in the interim. They just failed to agree with themselves.

Here is what each request is actually comparing you against.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

The Nine Pairings

Read the right-hand column again. Not one of those sources belongs to you. That’s the entire design.

Three of them do most of the damage, and they’re worth spelling out.

The IP one is structural, not administrative. A contractor who wrote 30% of the product in 2022 and never signed an assignment means the company does not own what it is selling. Remediation is possible, but it hands that contractor leverage at the precise moment you have none. Founders routinely spend six weeks locating someone who has since moved country, and term sheets expire inside six weeks.

The revenue bridge is where credibility dies. ARR in the deck, bookings in the model and recognised revenue in the accounts are three different measures, and every investor knows it. A gap is not read as a spreadsheet error. It is read as evidence that you present different numbers in different rooms, which is a character finding rather than a financial one.

The reference call is unarguable. “We’re seeing 95% retention” against “we’re not renewing” is the most common late-stage killer in venture, and there is no rebuttal available to you after it lands.

The One With No Cure

Everything above is fixable. It costs time, trust and negotiating position, but it’s fixable.

The 83(b) election is not.

Thirty calendar days from the stock being transferred. No extensions, no relief, no late filing. The Tax Court has rejected every argument put to it, including illness, attorney error and postal failure.

Three details cause most of the losses:

The clock starts at transfer, not signature - usually the date the board approved the grant, not the day you signed papers or the day shares appeared in Carta. Founders count from the wrong date and lose the window without noticing.

Filing with the IRS alone is invalid. A copy must also go to the issuing company. That’s a separate requirement, and written confirmation of receipt is what protects you years later.

“My lawyer handled it” is not a defence. It’s a personal filing. Verify it exists.

Since July 2025 there’s an electronic route via Form 15620 with immediate confirmation, now the sensible default. On paper, the certified mail receipt is your entire evidence base.

Check this tonight rather than at your next raise. A missing election gets more expensive every time your valuation moves.

Our Most-Read Articles Right Now

🔥 The Data Room That Doesn’t Lose Deals at the Last Minute Where to put everything, once you know what you’re proving.

🔥 Your Data Room Is Killing Your Round. Six Prompts to Fix It The structural errors that make commercial diligence read badly from page one.

📈 The Claude Due Diligence Playbook That Gets You to Closing Run the analyst’s first pass on your own company while you can still fix what it finds.

📋 What a Cap Table Actually Is, and Why It Matters The document diligence opens first, explained properly.

📋 The Most Dangerous Document a Founder Will Ever Sign The term sheet clauses that matter more than valuation.

📈 The Way VCs Actually Calculate Your Valuation The quiet maths that settles your price before the meeting starts.

The Window

Everyone says build a data room before you pitch. True, and insufficient, because a data room is somewhere to put documents and the failures above are disagreements between them.

The arithmetic of the calendar:

A priced seed round runs 12 to 16 weeks from first meeting to wire. Seed diligence takes two to four weeks; Series A runs four to eight, longer if you’re regulated. After the term sheet, you get roughly 30 days of confirmatory diligence.

That last month is when most founders exhale. It’s also when the deal still dies.

The cost of being caught out splits in two. The legal cleanup - cap table, assignments, consents - runs $5,000 to $25,000, and that’s the small number. The expensive part is the schedule. A contractor who has emigrated takes weeks to trace. Term sheets lapse while you look. Leads walk and don’t return.

The same work, run eight weeks before your first meeting, costs identical fees and none of the leverage.

Inside the Workbook

The download turns those nine pairings into something you can hand to a team and work through.

It opens on a Reconciliation Checklist - each area with its claim, its second source, the document that proves it, an owner, a deadline and a status flag, colour-coded so a glance shows where you’re exposed. Beneath that sits the Key Questions bank, each request rewritten the way a partner actually asks it, so you hear the question before they ask it.

Then the two that do the real work. The Three-Way Revenue Bridge puts deck ARR, model bookings and recognised revenue side by side and calculates the gap for you. The Cap Table Trace ties every share, SAFE, note and grant back to the board consent that authorised it, and flags whatever doesn’t match.

There’s also a Red Flag Register of the answers that should worry you when they come back, and an eight-week countdown mapped backwards from your first meeting.

Do These Five Things This Week

You don’t need the workbook to start. You need ninety minutes and a willingness to find something you’d rather not.

Confirm your 83(b). Ten minutes. The only item with no remedy.

Bridge your ARR. Open the deck, the model and the accounts together. Write down the gap. If you can’t explain it in one sentence, an investor won’t either.

Chase one signature. List everyone who has touched the product. Find the one name without an assignment on file. There is almost always one.

Call a customer you’re about to cite. Ask what they’d say if an investor rang. Better to hear it now.

Trace three lines of your cap table back to the consent that authorised them. If any don’t tie, assume the rest don’t either.

Where two sources disagree, either fix it or become the person who raises it. A disclosed problem is a negotiation. A discovered one is a repricing.

That’s the whole difference, and it costs you an afternoon.

Download My Due Diligence Workbook 👇

Everything above is yours: the nine pairings, the 83(b) trap with no cure, the timing that decides both. Act on that alone and you enter diligence ahead of most companies.

Below the paywall is the part that does the work for you.

✅ The Reconciliation Checklist. All nine areas as live rows: the claim, the source they check it against, the document that proves it, an owner, a deadline. Colour-coded, so one glance shows where you are exposed and who is fixing it.

✅ The Three-Way Revenue Bridge. Find your ARR gap before your investor does. Deck against model against accounts, difference calculated, reconciling lines itemised. Twenty minutes to rule out the most common cause of late collapse.

✅ The Cap Table Trace. Every share, SAFE, note and grant tied back to the consent that authorised it, with anything unmatched flagged. This is where diligence opens. It should be where you open too.

✅ The Key Questions Bank. All nine requests written the way a partner actually asks them, plus the follow-up when your first answer sounds thin. Hear the question before they ask it.

✅ The Red Flag Register. The answers that should worry you when they come back. Know which reference call went badly before your lead does.

✅ The Eight-Week Countdown. The whole thing sequenced backwards from your first meeting, so the work lands while it is still cheap and private.

Paid subscribers also get:

✅ 60+ additional tools across every stage of fundraising and company building: investor research, pitch deck screening, meeting prep, term sheet analysis, financial models, and the full investor database library.

Download Below

Start your 7-day free trial → Cancel anytime.

Read more

They Google “best startup accelerators 2026.”

They find a listicle ranking twenty programmes by brand.
They skim the cheque sizes and shortlist the three biggest.

Then they apply to whichever window is open and sign whatever arrives, because the name was familiar. Nobody ever tells them what the deal costs.

The problem isn’t the programme. The problem is that nobody prices the offer.

An accelerator taking a fixed percentage for a fixed amount of cash has already told you its valuation. It just never writes it down, and no listicle does the division for you.

So we did it. 255 times.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

Why the “Best Accelerators” List Doesn’t Work

Implied post-money = net cash ÷ fixed equity percentage

That’s the whole technique. Run it and the market becomes legible in ninety seconds.

a16z speedrun. $500,000 for 10%. Implied $5.0M.
Y Combinator. $125,000 for 7%. Implied $1.8M.
Entrepreneur First. $125,000 for 8%. Implied $1.6M.
Antler, Singapore. $150,000 for 10%. Implied $1.5M.
Techstars. $20,000 for 5%. Implied $400,000.

Same three months. Same demo day. A twelve-fold spread, and all of them publish the inputs while none publishes the number.

Read Techstars fairly, though. The 5% is common stock rather than preferred, putting them behind you in a downside exit. No fee. And most of their money sits in a $200,000 uncapped tranche that sets no valuation at all. A genuine defence, but not a reason to skip the calculation.

There’s a bigger problem with that league table, and it isn’t the one anyone expects.

The Second Number, and Why It Reverses the First

Implied post-money answers one question: at what valuation is this trade fair? Not the one you care about: what is this costing me, in dollars of my own company?

Cost = (your realistic valuation × fixed equity %) − net cash received

Say you could credibly raise at a $12M post-money cap.

Techstars takes 5%, worth $600,000, and gives you $20,000. Cost: $580,000.
a16z speedrun takes 10%, worth $1,200,000, and gives you $500,000. Cost: $700,000.
Y Combinator takes 7%, worth $840,000, and gives you $125,000. Cost: $715,000.
Antler Singapore takes 10%, worth $1,200,000, and gives you $150,000. Cost: $1,050,000.

On implied post-money, Techstars looked twelve times dearer than speedrun. On cost, Techstars is the cheapest of the four and speedrun is $120,000 more expensive.

The table doesn’t shift. It inverts.

The two numbers are linked exactly:

Cost = your fixed equity % × (your valuation − the programme’s implied valuation)

So implied post-money is your break-even. Below it, the programme pays you to attend. Above it, you pay them, and the bill rises in a straight line with how good you already are.

At a $4M valuation, speedrun’s $500,000 for 10% is worth more than the equity you give up. It effectively pays you $100,000 to be there. At $25M, the identical deal costs you $2,000,000. Same paperwork, same twelve weeks, a $2.1M swing.

Which is why the “is YC worth it?” argument never resolves. Both sides are right about different companies.

Caveat: the fixed slice converts inside your priced round and is then diluted by new money and any option pool increase, so true cost sits slightly below these figures. The ranking is unaffected.

The Evidence Nobody Puts in the Listicles

Gonzalez-Uribe and Leatherbee studied Start-Up Chile with a regression discontinuity design, in the Review of Financial Studies. Some founders got cash and coworking. Others got the same plus structured schooling.

The schooling significantly increased venture performance.

Cash and coworking alone showed no measurable effect whatsoever.

Sandy Yu, in Management Science, matched roughly 900 accelerator companies against 900 comparable ones and found they closed earlier and more often, consistent with accelerators resolving uncertainty about quality sooner. A 2025 meta-analysis of 21 studies found a positive effect overall, concentrated in top-ranked programmes.

Ranking by cheque size ranks by the weakest ingredient. The money buys runway, not performance, and per the formula above it’s the smaller half of the trade.

Brand tier isn’t snobbery, it’s the finding. A weak-network programme selling the same three months is selling the ingredient that didn’t work, at a price that still scales with your quality.

Our Most-Read Articles Right Now

🔥 What a Cap Table Actually Is - And Why It Matters What 7% actually becomes after three more rounds.

🔥 The Way VCs Actually Calculate Your Valuation The quiet maths that sets your valuation before the meeting starts.

📈 How Investors Decide If You Are Ready Stage by stage expectations, so you know what’s wanted next.

📋 The Most Dangerous Document a Founder Will Ever Sign The clauses that matter more than valuation.

The Fee That Removes a Quarter of Your Valuation

Several programmes charge a fee and deduct it from the investment, so you never write a cheque and never feel it leave.

500 Global’s published deal is $150,000 for 6%. Its own page states it charges $37,500, deducted from the investment. You receive $112,500.

On the headline: $150,000 ÷ 6% = $2.5M implied.
On what arrives: $112,500 ÷ 6% = $1.875M implied.

The fee removes $625,000 of implied valuation without appearing in the announcement.

Founder Institute sits at the far end: a fee reported between $499 and $999, plus 2.5% in warrants, and no cash at all. That’s tuition, and should be priced as tuition.

Y Combinator says the same on its own deal page: deduct fees from the investment when comparing offers. Always divide by net cash, never headline cash.

The Part of the Deal With No Price

Most deals arrive in two pieces, and only one is priced. The fixed slice is a set percentage for a set amount. That’s the price.

The uncapped MFN tranche converts on the terms of the lowest-cap instrument you issue before your next priced round. That isn’t a price. It’s a promise to accept whatever price you negotiate later, and an MFN adopts the most favourable terms in the window, so one small bridge safe taken in a difficult month drags the whole tranche down with it.

Note what this does to the cost formula. Cash in roughly equals value out at whatever cap you set, so the tranche’s net cost is near zero. The entire cost of an accelerator sits in the fixed slice. Which is why Techstars, with a small fixed percentage and a large uncapped tranche, is cheap for a strong company despite that alarming implied number.

What Each Entry Includes

1. Category. Seven products get sold under one word. Comparing across them is the commonest error.
2. Region and sector focus. Filter before reading anything else.
3. Headline cash, fee and net cash. The gap between the first and third is where the price hides.
4. Fixed equity percentage. The priced slice only, where your entire cost sits.
5. Implied post-money. Calculated live. Correct an input and it recalculates.
6. The uncapped tranche. Named, so you know which part has no price.
7. Length, cadence and acceptance rate.
8. Terms status. Verified at source, Published, or Not published, meaning they negotiate per company and you can’t compare in advance.
9. Direct link and checked date. Techstars changed its deal in 2025. Neo swapped its accelerator for a residency in 2026. Half the comparison content on Google describes deals that no longer exist.

A Taste of What’s Inside

a16z speedrun, San Francisco
$500,000 for 10% · Implied $5,000,000 · 12 weeks · Under 0.4% acceptance

The highest break-even of any major programme: close to free if you’re early, expensive if you’re not. A further $500,000 reserved for your next round. No board seat.

Techstars, 50+ cities
$20,000 for 5% common stock · Implied $400,000 · 3 months · No fee

The lowest implied price and the most misread. The fixed slice is small, so it’s the cheapest of the big programmes once you’re worth $10M or more. Spring 2027 closes 18 November.

Hub71, Abu Dhabi
No equity · Subsidy reported up to $100,000 · Up to 2 years

Housing, office space, insurance and credits, no ownership taken. Founders reject packages like this for “not coming with a cheque,” then sign an equity deal with less net cash in it.

The Geography Breakdown

274 entries, six continents.

US: SF, New York, Boston, Berkeley, Austin, across deep tech, health, space, defence and quantum.

UK and Europe: Seedcamp, EF, Zinc, Carbon13, Rockstart, Lanzadera, Venture Kick, the EIC Accelerator.

Gulf: Hub71, Flat6Labs across five countries, Sheraa, DIFC.

Africa: Baobab, Startupbootcamp AfriTech, Norrsken, CcHUB, MEST.

India and Asia: Antler India, EF Bangalore, 100X.VC, Surge, Iterative, SparkLabs.

Oceania and Canada: Startmate, Icehouse, CDL, DMZ, Velocity.

Latin America: Platanus, Parallel18, Start-Up Chile.

Geography moves the price more than category does, and the same brand prices differently by city. Antler, Flat6Labs, 500 Global and Startupbootcamp all run different deals per market.

The Windows Open Right Now

Techstars Spring 2027 opened 24 August and closes 18 November 2026. All Spring programmes share the deadline, so choose city and vertical first, then write once.

a16z speedrun SR008 has a priority window of 12 October to 1 November 2026.

Entrepreneur First’s first selection round closes 5 October 2026. Three run per cycle, and the first is strictly better than the third.

Y Combinator Winter 2027 isn’t published yet, but the pattern suggests late October. Early Decision is open now if you need an answer ahead of a raise or a visa.

“You are not buying capital. You are buying a deadline and a distribution list. Price them separately, because only one of those appears on the term sheet.”

The Bottom Line

Founders choose badly because they compare badly.

They rank by cheque size, the ingredient doing the least.
They ignore fees that remove a quarter of the implied valuation.
They treat an uncapped tranche as free money rather than a price their worst round sets later.
And they never ask the only question that matters: at my valuation, what does this cost me in dollars?

Two calculations fix all of it.

Implied post-money = net cash ÷ fixed equity. That’s your break-even.
Cost = fixed equity × (your valuation − that break-even). That’s your bill.

Run both before you write the application, not after the term sheet lands. If the bill is a number you wouldn’t spend in cash for the same outcome, you have your answer.

Techstars closes 18 November. Speedrun opens 12 October. Those applications are being decided right now, by whoever is doing the maths.

The file is waiting below.

The Full 255-Programme Directory

Exactly what you’re downloading:

• Directory. 187 equity programmes across six continents, with a live Implied Post-Money column, fee-adjusted net cash, uncapped tranches named, acceptance rates, source links and checked dates. Filterable by category, region, sector and stage

• Price Calculator. Enter the offer and your realistic cap. Returns the implied post-money, the fee-adjusted price, the MFN exposure, and the dollar cost at your own valuation

• Equity-Free & Credits. 39 programmes taking no ownership, from NVIDIA Inception and MassChallenge to Hub71 and the EIC Accelerator

• Peer Groups & Coaching. 22 options with entry bars and reported costs. The founder who can’t get into YC can usually get into one of these next month

• Events & Summits, Deadline Calendar, Summary and Sources. 26 events with typical timing, what’s open and closing next, live counts by region, and every figure cited and linked

Filter to your stage, rank by cost rather than cheque, then price your live offer before you sign anything.

Paid subscribers also get:

60+ additional tools across every stage of fundraising and company building: investor research, pitch deck screening, meeting prep, term sheet analysis, financial models, and the full investor database library.

Download Below

Start your 7-day free trial → Cancel anytime.

Read more

Let’s get into it.

In-Depth Insights 🔍

• Granola’s Growth Playbook: From $0 to $1.5B 📝
  Granola reached a $1.5B valuation in three years by delaying launch, cutting half its features, and avoiding the crowded meeting-bot category. It targeted VCs for dense feedback, used costly frontier models as a moat, and expanded into MCP and APIs as an agent context layer.

  

• Agent Memory: The Missing Layer for Persistent AI 🧠
  Agent memory lets systems retain useful state across sessions, since larger context windows still reset when a conversation ends. The core patterns are working, procedural, semantic, and episodic memory, with persistent stores using metadata to reliably scope information by user and type. [Developers Blog]

• Did the Cost of Starting a Startup Just Fall Off a Cliff? 💸
  Solo founders grew from 23.7% of startups in 2019 to 36.3% by mid-2025 as AI development costs fell roughly 100x in under two years. But funding has not followed: solo founders receive just 14.7% of VC dollars, while AI application companies often operate at only 0% to 30% gross margins.

  

• The One-Person Business, Built on Claude 💼
  Simple AI services such as missed-call text-back have little pricing power when established platforms already offer the same feature for $97 to $399 a month. The stronger opportunity is domain-specific integration and document work, with pricing tied to replacement cost and careful attention to German compliance rules.

  

• Why You Should Invest in the Most Expensive YC Startups 📊
  Among 642 YC companies from 2021 to 2024, those priced above their batch median reached Series A or later 25% of the time versus 8% for below-median companies. The same group had a 6% shutdown rate versus 16%, while median SAFE caps nearly tripled from $15M in W21 to $40M in P26. {Jared Heyman]

  

• The 3 New Pricing Models Replacing Seat-Based Pricing 💸
  Seat pricing is running out of room as software costs rise and 45% of CIOs now fund AI budgets from existing software spend. Consumption, outcome, and resolution pricing are emerging as alternatives, charging for usage, measurable business results, or discrete completed events. [SaaStr]

  

• Max Mullen on When to Take Advice From Investors 🥕
  Instacart co-founder Max Mullen divides investor advice into science, religion, and art, with founders relying on investors mainly when there is an objective answer. He also describes product-market fit as a progression, with Instacart taking four to five years to move from early adopters to weekly grocery shoppers. []

Social Media Gems 💎

• Dad’s 20s Starter Pack Explains Family Finances 😂
  A dad’s 20s included Patagonia, Zyn, LinkedIn, a road bike, an Apple Watch, and an office chair, apparently leaving little room for savings. The joke lands because the modern starter pack looks less like reckless spending and more like a monthly subscription to adulthood.

  

• Humanoid Robots Are Learning to Compete 🤖
  More than 2,000 humanoid robots from 666 teams competed at the World Humanoid Robot Games, a sharp jump from machines that needed tethers just two years ago. The next bottleneck is no longer whether they can move, but whether hardware costs can fall enough for factories, logistics, and elder care.

• The Entire Internet Runs on a Surprisingly Small Stack 🕹️
  A hand-drawn infrastructure stack maps modern web services across AWS, Cloudflare, DNS, open source developers, Microsoft, and now AI. The joke works because a huge share of the internet ultimately depends on a surprisingly compact set of infrastructure layers and unpaid maintenance.

  

• Founders Treat Doubt Like a Preflight Checklist 😅
  Most people worry about lacking skills, failing, or looking ridiculous, while founders often move ahead with the same doubts sitting in the cockpit. The difference is less about confidence and more about being willing to start before every answer is available.

  

Trending News ⚡

• AI Power Could End Up in Too Few Hands ⚠️
  Sam Altman warned that AI could eventually become concentrated among a small number of companies, models, or individuals. The core risk is that safety concerns could push society toward centralized control at the expense of individual choice. [Business Insider]

  

• Leaked Microsoft Data Shows the AI Talent Price War 💰
  An internal spreadsheet covering nearly 600 employees reveals salaries, raises, bonuses, and stock awards across Microsoft teams. The data highlights how compensation varies across Azure, Cloud + AI, and Microsoft AI as competition for talent intensifies. [Business Insider]

• Amazon Maps Out Fully Automated Package Sorting 📦
  Project Tetromino outlines plans for automated delivery stations, including a reported $103 million pilot targeted for 2028. Amazon disputes the figures, but the broader direction points toward reducing manual sorting across its logistics network. [Business Insider]

  

• Grafana Passes $600 Million ARR as AI Moves Into Production 📈
  Grafana Labs says customers are moving from AI experiments to production deployments, pushing annual recurring revenue beyond $600 million. The next challenge is operational, with companies increasingly focused on controlling costs, governance, and reliability at scale. [Business Insider]

  

• Anthropic Bets on a $30 Trillion Market Ahead of IPO 💰
  Anthropic is reportedly pitching investors on a total addressable market above $30 trillion and 2028 revenue near $190 billion to $200 billion. Those forecasts will matter heavily because the company’s IPO valuation depends on investors accepting a very large future growth curve. [Reuters]

• Meta Settlement Puts Global Pressure on Social Media Rules 🌐
  Meta’s potential $18 billion settlement with U.S. states over teen social media harms is drawing attention from regulators worldwide. Officials across Australia, Asia, Mexico, and Europe are now exploring whether similar protections should apply to young users. [Reuters]
  

• Nvidia Eyes Hugging Face at a $13 Billion Plus Valuation 🤝
  Nvidia is reportedly in talks to acquire Hugging Face at a valuation above $13 billion, adding another major AI platform to its orbit. Microsoft has also held discussions with Hugging Face, but Nvidia already has ties to the company through its 2023 investment. [Business Insider]

• SoftBank Targets Majority Stake in 1X at $6 Billion Valuation 🏭
  SoftBank is reportedly negotiating to buy a majority stake in humanoid robotics startup 1X Technologies at a valuation near $6 billion. The move fits its broader robotics strategy and follows its acquisition of ABB’s robotics business last year. [Reuters]

New Funds 💰

• Blackbird raised $750M across two venture funds targeting early-stage and growth-stage startups across Australia and New Zealand.

• Navisalma launched as a deep-tech venture firm and is seeking to raise €450M for its debut fund to back frontier technology companies.

• QuantumLight closed its second fund at €432M to back growth-stage companies across AI, fintech, SaaS, healthtech, and deep tech.

• Permanent Capital Ventures closed its second fund at $200M to back Series A applied-AI companies with a focus on scaling go-to-market execution.

• Cognition Venture Partners launched a new venture firm and is raising $170M to invest in cybersecurity startups serving the agentic AI era at Seed and Series A.

• MassMutual Ventures launched its $150M Climate Technology Fund II to back early-stage climate-tech and AI companies improving real assets across North America.

• FGV Capital closed Fund II at $35M to back early-stage companies at the intersection of fintech, financial access, and AI.

• Capital F closed its debut fund at $17M to invest at pre-seed and seed in technology companies serving the $15T female economy.

That’s a wrap for this week.

If you want to stay ahead of the curve, make sure you're subscribed. No fluff, just real startup insights delivered straight to your inbox.

See you next time 🚀

This series is a preview of a new book I’ve written with my partners Richard Blundell and Paul Watson — The Selling Software Algorithm: The Go-to-Market Navigation System for B2B Software Leaders.

It isn’t a strategy deck dressed up as a book. Between the three of us we’ve spent close to seven decades inside B2B software as founders, operators, investors and coaches — building companies, exiting some, and quietly breaking a few along the way, which is where most of the real learning lives. Through our work at Vencha we’ve now supported hundreds of founders trying to do the single hardest thing in software: turn a clever product into predictable commercial traction in a market that’s noisy, cautious, and more crowded than it has ever been.

One pattern shows up again and again. Companies rarely struggle because the product is bad. They struggle because they have no navigation system for going to market. This book is that system.

Get my Copy

Most founders think they know what a good customer looks like. The ones with the biggest problem. The ones who reply fastest. The logos that would look best on the website. The accounts that make the pipeline slide feel reassuringly full.

The problem is that need and accessibility are two completely different signals. A customer can desperately need what you sell and still take eighteen months to buy it. Another can book a demo tomorrow and have almost no reason to stick around once the novelty wears off. Treat both as equally good opportunities and you end up spending your best people on deals that were structurally wrong from the start.

There is a much simpler way to see it. Two axes. Four zones. Plot how badly an account needs you against how realistically you can reach, sell and deliver to them, and suddenly the market looks very different. The uncomfortable part is that your real target market is probably much smaller than the one in your pitch deck. That is exactly why this exercise works.

Table of Contents

• A Simple Example: Two Buildings on the Same Street

• Two Axes, Not One

• The Four Zones

• Why the Respond Zone Feels Like Validation

• Why the Research Zone Costs the Most

• Your Focus Zone Is Smaller Than You Think

• Filling In the Grid Honestly

• Matching Your Effort to the Grid

A Simple Example: Two Buildings on the Same Street

Picture two buildings on the same street. The first has its door propped open. Reception waves you through, the head of operations replies to emails within a day, and you could be in front of the person who signs inside a fortnight. The only problem is that nobody in there is losing sleep over the thing you fix.

The second building has security on the door, a procurement portal, and a fourteen-month cycle. But every person inside feels your problem daily, and has done for years.

Both look like opportunities on a Monday morning. Neither is where your next twelve months are won. Every account on your list sits somewhere between those two buildings, and the position is set by two things that have almost nothing to do with each other: how badly they need you, and how easily you can reach them.

Two Axes, Not One

Product Market Fit describes how well your product solves a viscerally painful problem for a given customer. It’s high where the pain is acute, persistent and visible, and where your solution removes it in a way that feels natural and complete. These are the organisations that recognise themselves immediately in your messaging and feel genuine relief when they see what you’ve built.

Go to Market Fit isn’t about the product at all. It’s about access and efficiency: how easily you can identify the right personas, reach them through credible channels, hold a meaningful conversation, and carry them through a buying and onboarding process that matches how they actually operate. High Go to Market Fit also tends to mean manageable integrations, proportionate procurement, accessible budget, and early value that arrives fast enough to create advocacy and referrals.

Most founders only ever score the first axis. That’s why the “is this a good account?” conversation goes round in circles. Two people can both be right and still be arguing, because one is talking about pain and the other is talking about access.

Plot both, and the fog lifts. What was a subjective debate about good and bad opportunities becomes an evidence-led map that shows you where to invest and where to step away.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

The Four Zones

Top right, where both are high, is your Focus zone. These customers feel the pain deeply and you can reach them easily. Sales cycles are shorter, implementations are cleaner, renewals are more predictable. This is where momentum lives, and where nearly all your time, attention and resource should go in the early stages.

Bottom right is the Respond zone. Easy to reach, poorly aligned. They come inbound or answer outreach quickly, and it’s tempting to treat that speed as validation.

Bottom left is the Ignore zone. Low fit, low access. They neither need your solution nor are reachable, and a sale here is churn waiting to happen. If you’ve done the Perfect Customer Profile work properly, very little of your list should sit here, but it’s worth naming so the team knows what it looks like.

Top left is the Research zone, and it’s the subtle one. Product Market Fit is high, sometimes enormously so, but reaching the Economic Buyer is hard and the path to purchase is long. Think Walmart in the US, or the NHS in the UK.

Why the Respond Zone Feels Like Validation

Inbound feels like proof. Someone found you without being chased. Someone replied on the first email. Someone booked a call and turned up.

But the signal you’re reading is about your own accessibility, not about their pain. A fast reply tells you the channel worked; it tells you nothing about whether doing nothing is unacceptable for them.

There’s nothing wrong with responding professionally and helpfully to these accounts. The mistake is letting them shape your strategy. They’re opportunistic rather than repeatable, and pursued too aggressively they’ll quietly pull your roadmap, your messaging and your best people away from the customers who actually matter.

If you’re building right now, you’ll need these:

• 243 Ways to Fund Your Startup Without Giving Up a Share: 243 programmes across 63 regions, and the eligibility gate that quietly disqualifies most applicants.

• 15 Claude Skills That Run Your Entire Raise: install once, and your story stops drifting somewhere around investor number forty.

• The Investors Who Are Actually Writing Cheques In H2: 287 family offices your warm intro list doesn’t know exist.

• The Claude Guide Every Founder Should Run Before Fundraising: nine prompts that replicate the screen your deck meets before a partner ever opens it.

• How to Build Your Fundraising Narrative with Claude: the five-prompt sequence that puts your story in the order it needs to land, not the order it happened to you.

Why the Research Zone Costs the Most

This is the more expensive mistake, and it’s usually made by the smartest founders in the room. The logic is hard to argue with: the pain is real, the account is enormous, and landing one would change everything.

It might. But decision cycles stretch over many months or years, procurement is complex, and adoption is slow. The quadrant holds genuine future potential, and it’s worth researching, nurturing and learning from. It simply cannot carry your near-term growth. Place your future there too early and you’ll run out of time before you run out of ideas.

Your Focus Zone Is Smaller Than You Think

When teams fill this grid in honestly, the Focus quadrant is almost always smaller than expected. That’s normal, and it’s healthy. Software businesses don’t grow by pursuing breadth; they grow by exhausting depth.

Seeing your accounts distributed across the grid is usually sobering. It exposes how much energy went into deals that felt exciting at the time but were structurally misaligned with how the business needs to scale. That isn’t a reason for regret. It’s evidence of learning, and the point of the exercise is better decisions ahead, not judgement of the ones behind.

Filling In the Grid Honestly

Get the room right first. Sales, Marketing, Product, Tech and Customer Success, phones and laptops off. The diversity of perspective isn’t a risk here, it’s the whole point, and the most useful learning tends to surface in the disagreements.

Ask sales which accounts made it easy to get to the Economic Buyer, and which felt disproportionately hard work. That’s your Go to Market Fit axis, described by the people who lived it. Ask Customer Success where the product delivers value most naturally and where friction persists despite everyone’s best efforts. Ask Marketing which segments respond and which stay stubbornly indifferent.

Bring data where you have it, such as cycle length, onboarding time, product usage and renewal behaviour, but give equal weight to lived experience. The stories your team tells about customer interactions are signals, not anecdotes.

Then capture not only where each account sits, but why. The reasons matter more than the position, and they’re what make the grid useful again in six months when your product and your reputation have moved.

Matching Your Effort to the Grid

None of this is static. Revisit it quarterly and watch the points move as the product matures and new routes to market open up. Success isn’t filling the grid evenly. It’s making the Focus quadrant dense, deliberate and unmistakable.

This is exactly what Builder Step 6 exists to produce: your market map, sorted into focus zones, so you can see at a glance whether your team’s effort is landing where fit is strong and access is real, or drifting into the two quadrants that feel productive and aren’t.

Spend your life where fit is strong and access is real. Everything else can wait.

Selling Software Algorithm Builder, Step 6: Market Map (recreated illustrative mockup based on the tool’s documented output; a fictional example company is shown)

Next in the series, we look at what happens when the algorithm stops being a document and becomes behaviour, and why this is the one piece of work a founder cannot delegate.

-Chris Tottman

Most account lists get sorted by company size or logo appeal, rather than by whether you can actually reach the person who signs.

The Selling Software Algorithm Builder Step 6 generates your market map and sorts your accounts into focus zones, so you can see where your effort is really landing.

Start your 7-day free trial → Cancel anytime.

Build Your Market Map

Read more

They Google “startup grants 2026.” They find a listicle with twelve entries, eight of which are American. They click three links. Two are dead. One leads to a programme that closed in 2023.

They conclude the whole category is a waste of time and go back to building the deck.

Meanwhile, US federal innovation programmes alone deploy over $4 billion a year into small companies. The European Innovation Council has €1.4 billion committed for 2026. Innovate UK, NRC IRAP, Bpifrance, CDTI, Enterprise Singapore and the Israel Innovation Authority add billions more.

None of it takes equity. None of it takes a board seat. None of it appears on your cap table.

The problem isn’t that the money doesn’t exist. The problem is that almost nobody tells you what actually disqualifies you.

We’ve spent weeks fixing that. The result is a directory of 243 non-dilutive programmes across 63 regions, every one checked against source in August 2026. Award sizes, cadences, realistic odds, direct apply links, and the specific eligibility gate that silently rules most applicants out before anyone reads a word of their submission.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

Why Grant Applications Actually Fail

Not because the idea was weak. Because founders write grant applications the way they write pitch decks.

Those are opposite skills.

When you pitch an investor, you are selling a distribution of outcomes. Your job is to make the top of that distribution feel plausible. You lead with the best number. You compress the risk into one slide nobody reads properly. Optimism is the product.

A grant assessor is doing something completely different.

Take Innovate UK as the clearest published example. Your application goes to three to five independent assessors. Each scores individual questions out of 10 against a rubric published before you apply. Scores are moderated across the whole round. Success rates sit in the 10% to 25% band, lower in oversubscribed competitions.

Which means the realistic funding bar is a mid-80s percentage, not the 70 you were expecting.

Assessors are not hunting for reasons to fund you. They are hunting for reasons to deduct.

Three things follow from that, and each one inverts a habit that serves you well in a partner meeting.

1. Unhedged optimism is a deduction, not a strength. Innovation bodies exist to fund the removal of technical uncertainty. If your application reads like there is no meaningful risk left, you have just argued that you don’t need the grant. Name the hard unknown. Then show a credible plan to resolve it.

2. Every sub-clause is a scoring line. Grant questions are three or four questions stapled together with “and.” Founders answer the interesting part, run out of word count, and skip the last clause. That clause carried marks.

3. Scope is checked before merit. Eligibility and scope screening happens before an assessor reads a word of your technical case. A brilliant project in the wrong competition dies without anyone ever reading it.

Stop pitching. Start scoring.

Grants Are Five Different Products Wearing One Word

This is the second reason founders write the category off.

They hear “grant,” picture one thing, and go straight to the hardest version of it.

Layer 1: Credits. Infrastructure and tooling burn. $1k to $350k in kind. Live in days. Near certain at entry tier.

Layer 2: Fellowships and micro-grants. Funds the founder personally, not the company. $500 to $250k. Two to twelve weeks. Assesses you, not your traction.

Layer 3: Tax relief. Pays you back for work you already did. 15p to 27p per £1 spent. One accounting cycle. Near certain if eligible.

Layer 4: Innovation grants. A defined R&D project. $50k to £2m. Four to nine months. Scored against a published rubric.

Layer 5: Blended deep tech. Scale-up and market entry. €2.5m grant plus optional equity. Six to twelve months. Around 3%.

Almost every founder who “looks into grants” goes straight to Layer 4. It is simultaneously the slowest, the most competitive and the most writing-intensive of the five.

Layers 1 and 3 are close to guaranteed money for a company that already qualifies, at a fraction of the effort.

Sequenced properly, each layer buys the runway to work on the next one.

What Each Entry Includes

Every one of the 243 entries gives you nine things:

1. Award size. The real number, not the marketing ceiling. Where a headline figure is only reachable via VC nomination, we say so.

2. Layer. Which of the five above, so you can sort by effort rather than by size.

3. Region. 63 of them, from US state match programmes to Indian state schemes.

4. Stage fit. Pre-incorporation, pre-revenue, post-Phase II, scale-up. Applying at the wrong stage is the most common wasted application.

5. Cadence or next deadline. Monthly, quarterly, annual, rolling, or a specific 2026 date.

6. Effort. Low, medium, high, very high. A Layer 1 credit is an afternoon. An EIC Accelerator bid is a campaign.

7. Realistic odds. Drawn from published success rates where funders publish them. A planning aid for ranking your effort, not a promise.

8. The eligibility gate. The single most valuable field, and the one no other list has. Written as the specific thing that disqualifies people, not as generic criteria.

9. Data confidence. Verified against source, directional, or amount unpublished. So you know which figures to double-check before you build a plan around them.

Our Most-Read Articles Right Now

🔥 You get one meeting. Here’s how to prepare for it in 20 minutes with Claude. Most founders spend their prep time on the pitch. The founders who close rounds spend it on the investor.

🔥 The 300+ AI Investors Actually Worth Your Time 305 named individuals, not funds. LinkedIn URLs, real theses, and a specific warm intro playbook for each one.

🔥 I asked Claude to reject my pitch until it couldn’t. Then I raised. The preparation system founders are using to walk into VC meetings already knowing every question that kills deals.

📈 The Number That Kills More Fundraises Than Any Bad Idea Every investor asks it. Most founders get it wrong. Why your market size slide is either your strongest asset or the quiet reason you’re not getting a second meeting.

📋 The 8 Checks That Can Kill a Signed Term Sheet 15% to 25% of signed term sheets never make it to a closed round. Not because the business got worse.

📋 The Due Diligence Playbook: What VCs Actually Analyse Before They Invest An inside look at how venture capital firms evaluate founders, financials, structure and scalability before committing capital.

A Taste of What’s Inside

Four entries from across the directory.

Enhanced R&D Intensive Support (ERIS), HMRC United Kingdom · Layer 3 · Up to 26.97p per £1, paid in cash

For loss-making SMEs whose R&D spend is at least 30% of total expenditure. No pitch, no assessor, no competition. A pre-revenue UK startup spending £400,000 a year on qualifying development is looking at a cash credit near £100,000. The gate: first-time claimants must notify HMRC within six months of the end of the period of account. Miss it and an otherwise valid claim is void.

NSF SBIR/STTR, National Science Foundation United States · Layer 4 · Up to $305,000 Phase I, $1.25m Phase II

Deep tech, any field of science and engineering. NSF takes no equity and you keep your IP. The gate: you cannot submit a full proposal at all until you file a three-page Project Pitch and are formally invited. Processing takes around three weeks and you are limited to two pitches per submission window. Burn both on weak pitches and you are locked out until the next window.

Amber Grant, WomensNet US and Canada · Layer 2 · $10,000 monthly, plus larger year-end awards

At least 50% women-owned. Short application, small fee. The quiet advantage: because it runs monthly, you compete against one month of applicants rather than a year’s worth. That makes the odds structurally better than almost anything else in the directory.

i-Lab (Concours d’innovation), Bpifrance France · Layer 4 · Non-repayable subsidy up to €600,000, no equity taken

One of Europe’s most prestigious deep tech awards. The gates: a French-incorporated company (founders of any nationality), a genuinely scientific or engineering core, and applications are expected in French. Software-only ideas rarely qualify. Around a 5% full-award rate, so treat it as a campaign, not an opportunistic bid.

Four completely different shapes of money: a near-certain tax credit, a gated federal grant, a monthly micro-grant with unusually good odds, and a competitive European deep tech award.

The directory covers all of it, and the 239 other programmes across every layer, stage and geography.

The Geography Breakdown

243 programmes. 63 regions.

United States. All eleven SBIR agencies plus ARPA-E and ARPA-H, NSF I-Corps, and the state match programmes that turn a $295,000 federal award into $445,000 if you happen to be incorporated in the right place.

United Kingdom. Innovate UK Smart Grants, Women in Innovation, KTPs and Innovation Loans, plus all three devolved nations. Scottish Enterprise, Business Wales, Invest NI. Growth Hubs and council-level schemes that decide in two to four weeks.

Europe. The full EIC suite and Horizon Europe clusters, Eurostars, plus the national programmes that almost no English-language list covers properly: EXIST, ZIM and the German Research Allowance, Bpifrance’s ladder from Bourse French Tech to i-Lab to ADD, Spain’s NEOTEC and CDTI, the Dutch WBSO and MIT, Enterprise Ireland, Vinnova, Innovation Norway and Business Finland.

Rest of world. Canada’s four IRAP streams plus SR&ED and the provincial layer. Singapore, India’s central and state stack, Israel, Australia, New Zealand, Hong Kong, the UAE, and pan-African programmes.

What to Do This Week

Claim your Layer 1 credits. An afternoon of forms. Start with NVIDIA Inception, which is free, open globally at any stage, and materially improves your standing with the hyperscaler programmes. Then build a spreadsheet with one column for expiry date. Founders routinely let five figures of credit lapse because nobody owned the calendar.

Start any registration that gates everything else. SAM.gov takes three to six weeks and blocks every US federal application until it is active. DPIIT recognition in India takes two to three weeks and gates almost every central scheme. Both are free, and neither requires you to have chosen a programme yet.

Book an hour with your accountant about Layer 3. If you are UK-based and doing genuine technical work, this is almost certainly the largest predictable non-dilutive cheque available to you, and most founders treat it as year-end admin rather than a financing decision with a hard deadline.

Then, and only then, shortlist four Layer 4 programmes. A grant application takes 40 to 200 hours. Four is a campaign. Twelve is a fantasy.

The Traps That Kill Applications Before Anyone Reads Them

Almost none of these are about the quality of your idea.

You started too late. Registrations, consortium formation and national co-filings eat weeks before you can submit anything. A Eurostars proposal takes four to eight weeks to prepare, and the national documentation is due within days after the deadline. Founders lose to the calendar far more often than to the competition.

You spent money you were about to be given. Canada’s NRC IRAP agrees funding before you incur the expense. Costs already spent are disqualified. Applying after the work is done wastes the entire application.

You burned an attempt on a draft. Rounds are finite and several programmes cap resubmissions. Submitting something that scores in the low 70s is not a free option. It costs an attempt, six to twelve months of waiting, and it attaches your name to a weak application.

You applied to something that no longer exists. Seven major programmes still appearing on widely-circulated 2026 lists have closed or been renamed. SDTC folded into IRAP. Callaghan Innovation was disestablished. Accelerating Commercialisation closed in 2023. EIT Manufacturing shut in April 2026. Every one of those is still being recommended somewhere.

One Rule Before You Apply to Anything

Print the published assessment criteria and put them next to your draft.

Map every paragraph you write to a specific scoring line. Where the criteria use a particular word, use that word instead of your more elegant synonym.

Assessors are reading forty applications in a fortnight. They are scanning for the thing they were told to look for.

“The best founders don’t spray and pray. They submit four applications that matter. That takes knowing exactly which four you’re actually eligible for, before you write a single word.”

The Bottom Line

Founders don’t miss grant money because they build badly. They miss it because they search badly.

They read a list that’s shallow, stale and 80% American. They go straight to Layer 4, the slowest and hardest of the five. They write the application like a pitch deck. Then they find the disqualifying eligibility rule in week three of drafting instead of week zero.

And the clock is always further along than it looks. Three major windows sit inside the next two weeks: the EIC Accelerator cut-off on 2 September, the NIH standard due date on 5 September, and Eurostars Call 11 closing 10 September. Every one has a lead time in front of it longer than the time remaining.

Which is exactly the point. The next cut-offs are 4 November and 7 January, and whether you make those is being decided this week by whether your registrations are already running.

243 programmes. 63 regions. Every eligibility gate written out. Every figure dated.

What One Row is Worth

One ERIS claim on £400,000 of qualifying development returns roughly £100,000 in cash. No pitch. No assessor. No competition.

One state match turns a $295,000 federal award into $445,000.

One missed HMRC notification voids the whole claim, however much qualifying R&D you actually did.

A Premium subscription costs less than an hour of the adviser you would otherwise pay to find any of it.

SAM.gov takes three to six weeks. EIC Step 1 takes four to six. Neither requires you to have picked a programme yet, and the file tells you which to start today.

📥 The Full 243-Programme Directory

Seven tabs, built to be used in twenty minutes rather than read for an hour.

✓ Find Yours. Start here. A region-by-layer matrix showing exactly how many programmes exist for you and which to open first. A UK founder sees 28. A Canadian sees 21. Ten seconds to know where to look.

✓ The Directory. All 243, filterable by region, layer, stage and effort. Award size, cadence, realistic odds, direct apply link. Plus the eligibility gate on every single row, written as the specific thing that disqualifies people rather than as generic criteria.

✓ A built-in tracker. Mark each programme Shortlisted, Applying, Submitted or Won from a dropdown, and your progress rolls up to the front page automatically. This is a working file you reopen weekly, not a PDF you read once.

✓ Deadline Calendar. Every major cycle through to Q2 2027, with the registration or pre-gate lead time you have to clear first, and therefore your true start date.

✓ Stacking Rules. Which instruments legally combine and which cancel each other out. The US stack that turns a $295,000 federal award into $445,000. The European ceiling that quietly shrinks a later, larger award if you’ve already drawn against it. And the six conditions that void an otherwise valid claim.

✓ Assessor Scorecard. Score your draft across the eight dimensions a panel actually marks. It returns your weighted percentage against the mid-80s bar and tells you plainly whether to submit. Score below it and you’ve just saved an attempt and six to twelve months.

✓ Sources and Caveats. Every source named, every figure dated. Plus the seven programmes still circulating on 2026 lists that have quietly closed or been renamed, so you don’t lose a month applying to one that no longer exists.

Download Below

Start your 7-day free trial → Cancel anytime.

Read more

It is not 1 September.

It is this week. Right now, while partners are still away, half clearing an inbox on a phone, deciding which four or five names they want to look at properly when they get back.

That list is being written today. Quietly, informally, mostly in a notes app. And it is the single most important document in your fundraise, because it decides who gets the good half of September and who gets the polite half.

Founders treat the autumn as a starting gun. Investors treat it as a shortlist they began building in July.

After 20 plus years investing, and having raised as a founder myself, I can tell you the pattern is astonishingly consistent. The rounds that close before Christmas were almost never introduced in the autumn. They were introduced before it.

Here is how the season actually works, why 1 September is already nine days too late, and the four things to have in your hands before tomorrow week.

The Season Nobody Puts In Writing

Venture has two live seasons. September to early December, and January to June. Everything else is administration.

The autumn season has a shape, and it is not evenly distributed.

September is the flood. Partners return with a refreshed pipeline mandate and an empty diary. Six to ten decks a day. Meetings are cheap and easy to get. Almost nobody receives a term sheet in September. What actually happens is that each partner quietly builds a shortlist of eight to twelve companies they intend to keep watching.

October is the decision. Second meetings. Partner meeting airtime. Reference calls you will never hear about. By the end of October, most partners have privately chosen the one or two deals they are prepared to fight for internally before the year ends.

November is the paperwork. Diligence, investment committee, terms, legals. A partner is not shopping in November. They are defending a choice they already made in October.

December is the door. Deals close in the first week. After that the industry goes quiet until the second week of January, and January restarts the entire cycle with a fresh mandate and no memory of your November email.

An investor falls in love in October and is one and done. Everything after that is administration on a decision that was already taken.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

Why The Window Is Twelve Weeks, Not Four Months

Founders count September, October, November, December and see four months of runway. It is not four months.

From Tuesday 1 September to Friday 4 December there are 94 days. That is 13.4 weeks before you take anything out.

Then take things out.

• US Thanksgiving falls on Thursday 26 November. That entire week is dead for any fund with a US partner, a US LP base, or a likely US co investor. Not one lost day. One lost fortnight of momentum, because the week before becomes a scramble and the week after becomes a recovery.

• The first week of December is a closing week, not an opening week. Partners are pushing signed deals over the line before year end. A new name arriving that week gets filed, not read.

• Investment committees meet weekly at best, often fortnightly. Between 1 September and the last usable IC before Christmas you have roughly ten to twelve committee dates. Your deal needs two or three of them.

Strip out the dead weeks and the honest working window runs from 1 September to around 25 November. Twelve weeks.

Now hold that against the data.

In DocSend’s seed research, half of all successful raises took 13 to 24 weeks. Founders contacted an average of 66 investors and took 38 meetings to get there.

The autumn window is shorter than the average successful raise.

That is the whole article in one line. Founders who “start in September” are not starting early. They are arriving on the day the race begins with the preparation still to do.

Do this before you read any further. Open a calendar. Put your target close date on it. Count back twelve weeks. Whatever date that lands on is the day your raise actually had to begin, and for most founders reading this it is a date that has already passed.

The Slot Problem

This is the part that changes behaviour, so sit with it for a moment.

A partner at a traditional Series A or B firm does one to two new deals a year. At pre seed funds it might stretch to three or four. This is structural, not a market condition. Partners slow down at seven to ten board seats, and every cheque has to be plausibly capable of returning the fund.

Two live seasons a year. One to two deals a year.

Most partners have roughly one slot to fill this autumn.

Here is the funnel that produces it:

• 500 to 800 decks screened to generate 50 to 100 real conversations

• Around 20 opportunities taken seriously enough to warrant proper work

• One to two cheques

You are not competing to be good. You are competing to be the one.

And then there is the line that matters most. According to Harvard Business Review’s analysis of 900 venture capitalists, more than 70% of deals originate inside the firm’s existing network.

Read that carefully. The majority of funded companies were already known to the investor before a deck was ever formally sent. Not warm introduced on the day. Known. Seen twice. Mentioned by a credible founder in July. Followed since the spring.

When I was raising, I assumed the process began when I hit send. It does not. By the time you hit send, most of the decision architecture is already in place.

The founders who get funded in October were introduced in August.

What A November First Meeting Actually Means

Imagine you do everything right, just late.

You send a genuinely excellent cold email on 4 November. A partner replies. You get thirty minutes on 12 November. You are good in the room. They like you.

Then they open their own calendar.

To back you, they need a second meeting, which lands in Thanksgiving week, so realistically the first week of December. They need reference calls, so mid December, when half your customers are on leave. They need an IC slot, and the December committees are already full of deals that started in September. And their one slot for the year is either spent or spoken for.

So they say the most efficient sentence available to them:

“Really enjoyed this. Let’s stay close and pick it up in the new year.”

That is not a soft no. It is a real yes, delivered into a calendar with no room left in it.

A first meeting after roughly 5 November is a January meeting wearing a November coat.

And you do not get those names back. A partner who met you in November and said “let’s stay close” is far harder to re engage in February than a partner who has never met you at all. You have spent the introduction and received nothing for it.

Two Things Founders Get Wrong About Timing

Both of these are expensive, and I see them every single autumn.

Runway decides when you start, instead of the calendar

Most founders begin raising when the bank balance starts to frighten them.

That is the worst possible trigger, because fear is visible and investors price it. The calendar should decide your start date, and your runway should have been extended to meet it. If you need money in the bank by December, the decision to raise had to be made in the summer.

Founders quit six weeks in, right before it works

DocSend found something brutal in their data. Founders whose raises failed gave up after an average of 6.7 weeks. The successful rounds took twelve or more.

Six point seven weeks from 1 September is mid October. That is precisely the moment conviction is forming on the other side of the table, and precisely the moment silence feels most like rejection. A compressed season makes this worse, not better, because there is more silence packed into every week.

The founders who survive that fortnight are the ones who walked in with the map already drawn, and knew the silence was scheduled.

Where Are You, Honestly?

Four checks. Answer them properly, not generously.

• Could you send someone a written list of named partners, not firms, today?

• Has anyone credible mentioned your company to an investor in the last six weeks?

• If a partner asked for your data room on 3 September, could you send a link that afternoon?

• Do you know exactly what you are raising, at what price, and why that price survives their model?

Four yeses and you are in the small minority for whom September will be genuinely good. Go and enjoy it.

One or two and you have nine days, and nine days is enough, but only if you spend them on the right four things.

Zero and this autumn is your preparation season, and January is your raise. That is not a failure. It is the single most valuable decision available to you this month, and it is worth more than any amount of deck editing.

The point of the next nine days is to find out which of those three founders you are, before an investor finds out for you.

The Four Questions You Have To Answer Before 1 September

Every autumn raise resolves into four questions. Founders usually answer them in the wrong order, and usually by burning real meetings to find out.

The season is twelve weeks. You cannot afford to learn any of these from a partner.

1. Am I actually fundable at the round I am going for?

Not “do I have a story.” Do the signals match the stage.

Investors rarely reject a company because a single metric is off. They reject because the pattern does not match the stage. ARR fine, growth slow, margins thin, efficiency weak resolves into one sentence in a partner’s head, and the sentence is “not ready yet.”

The most expensive mistake in venture is starting a raise six months early. In a twelve week window it is fatal, because you spend the season proving something the numbers were never going to support, and you arrive in January having already been passed on by the people you most wanted.

Answer this before you write a single email.

2. Will my deck survive the first thirty seconds?

Investors spend roughly three and a half minutes on a deck, and forward two to four slides internally.

Those slides are the real product. Not the deck. The two slides a partner pastes into a channel with a one line note. If nothing in your deck is forwardable, your deal cannot travel inside the firm without you, and a deal that cannot travel cannot reach a committee, which means it cannot close in December.

In September a partner is screening six to ten of these a day. Thirty seconds is generous.

3. Which investors can actually say yes to me?

This is the quiet filter, and it is where most of the twelve weeks gets spent for nothing.

Stage, sector, cheque size, geography. Four gates, all invisible from the outside, all absolute. If your raise sits outside a fund’s band, the conversation was over before it started, and nobody will ever tell you that. They will simply be slow.

Founders contact an average of 66 investors to close a seed round. In a twelve week window, the difference between 66 correctly chosen names and 66 hopeful ones is the difference between closing in November and starting again in January.

4. What is the number, and does it survive their model?

The question arrives in every second meeting, usually in October.

“So, what valuation are you raising at?”

Most founders guess, or anchor high and hope confidence carries it. Investors are not guessing. They are running a return based model, working backwards from an exit value to the ownership they need, then pricing your round to produce it.

If you do not know that model, you are negotiating blind in the exact month the decision is being taken.

Why These Four, And Why Now

Individually, these are useful in any month of the year.

Together, in the next nine days, they are the whole season.

Answer question one and you know whether you are raising this autumn or preparing for January. Answer two and your deal can travel inside a firm without you in the room. Answer three and your twelve weeks are spent only on people capable of writing the cheque. Answer four and you walk into October pricing the round the way the person across the table already is.

Fundable. Forwardable. Targeted. Priced.

That is the entire autumn, and every one of those four answers is available to you before Tuesday week.

The Four Tools

All four are available exclusively to paid subscribers of The Founders Corner®, and all four are below.

✅ Funding Milestone Navigator - enter your metrics and it scores your readiness against real stage expectations, then hands you the priority gaps to close and what “good” actually looks like at your round. This is your go or no go on the autumn.

✅ The 30 Second Pitch Test - the fast filter every investor runs before deciding whether your deck earns a real meeting. Run your deck through it this week, not in October when it is being forwarded around a firm without you there to explain it.

✅ The Investor Fit Finder - a curated shortlist drawn from 2500+ continuously updated investors, matched to your stage, sector, geography and cheque size, with a transparent fit score on every name. Ten to fifteen right names instead of two hundred hopeful ones.

✅ The VC Method Calculator - the quiet maths behind your valuation, fully editable in Excel. Exit value, required ownership, dilution, IRR and MOIC, so the number you say out loud in October is one that survives their model.

…plus 60+ additional tools across every stage of fundraising and company building: investor research, pitch deck screening, meeting prep, term sheet analysis, financial models, and the full investor database library.

Start your 7-day free trial → Cancel anytime.

Get All Four Below 👇

Twelve weeks. One slot per partner. Several thousand founders reaching for it.

The window opens in nine days, and it does not open twice.

If you want to close a round in 2026, you’ll need these.

Read more

Let’s get into it.

First, something free 👇

The founders winning in 2026 are everywhere: LinkedIn, newsletters, podcasts, investor DMs. They run on a system that publishes for them, rather than carving out time they rarely have.

We built that system into a free AI Creator Kit: one Notion workspace with AI agents, templates, a creator CRM, and partner credits worth thousands.

Grab it here, no paywall:

Free AI Creator Kit

In-Depth Insights 🔍

• Top 20 VC Matches in 30 Seconds 🎯
  A free tool screens 293 active VCs, angels, and family offices by stage, sector, geography, check size, and thesis to surface relevant investors fast. Founders pitching at VC Pitch Conf on September 17 also get 20 guaranteed seven-minute meetings matched from the same investor pool.

  

• Delay Your First Engineering Hire 🧑‍💻
  AI-assisted development is pushing the point where startups need their next engineer, with some teams reaching $1M ARR using only one or two technical hires. The better hiring playbook starts with trusted networks and evaluates candidates on ownership and AI-assisted reasoning instead of traditional whiteboard performance.

  

• A Nine-Prompt Stack for Closing Funding Rounds 🤝
  The system covers investor research, objection handling, data-room analysis, and term-sheet review, beginning with context that frames Claude as a skeptical senior partner. The prompts are designed to pressure-test fundraising decisions, though the full stack and supporting libraries sit behind a paid subscription.

  

• Build AI Systems That Improve Themselves 🔄
  A nine-part framework combines goals, evaluations, memory, guardrails, escalation, and stopping conditions to turn isolated prompts into repeatable feedback loops. The Self-Improving Champion Loop tests changes against holdout data before deployment, alongside three other patterns, five failure modes, and a practical build specification.

  

• The NRR-GRR Gap Reveals Hidden Churn 📊
  NRR includes expansion while GRR excludes it, making the difference a useful signal for whether growth reflects broad customer retention or a handful of expanding accounts masking churn. The framework breaks SaaS businesses into four segments and highlights dunning improvements that can recover 70% to 80% of failed-payment churn. []

  

• Investors Are Losing Faith in Startup Metrics 🤔
  Fifteen VCs flagged ARR as especially unreliable because companies increasingly use it for run-rate, GMV, and consumption revenue without consistent definitions. CARR, gross margin, POC-driven NRR, and physical AI autonomy rates also drew scrutiny when underlying costs or ideal testing conditions are excluded. []

  

• From $8M ARR to Rebuilding From Scratch 🛍️
  Ruslan Leteyski built Checkout X to $8M ARR before a Shopify platform change eliminated the product overnight, forcing a complete reset. His new company Zipchat sells AI-powered sales agents to e-commerce stores through usage-based pricing, now processing 700K conversations monthly while growing 10% week over week. [Indie Hackers]

Social Media Gems 💎

• Thrive Capital’s Fund Returns Show the Exit Gap 📊
  Net IRRs range from 12% to 17.5%, while DPI stays below 0.3x for every fund younger than 10 years. The portfolio spans Instagram, Stripe, OpenAI, SpaceX and Anduril, giving each vintage a different path to realized returns. [Restructuring]

  

• Worst Value Cities: London, New York, LA and San Francisco 🌆
  A 100-city comparison finds these markets combine high living and housing costs with weaker quality-of-life value. Porto, Valencia and Prague have led the value rankings for seven years, while costs are rising fastest in several migration-heavy cities.

  

• FatFIRE’s Case for Buying Back Your Time 💰
  The argument is that once housing and core expenses are covered, additional wealth delivers smaller gains in day-to-day quality of life. Financial independence matters most when it creates room for health, family, good food and the freedom to spend time outside the balance sheet.

  

• The 2013 Excel File Still Has a Job 📊
  A meme captures a familiar enterprise reality: AI can replace workflows faster than companies can replace the spreadsheets built around them.

  

Trending News ⚡

• Stripe to Acquire OpenRouter for $7B+ 💳
  Stripe is set to acquire AI gateway OpenRouter for more than $7 billion, giving it infrastructure that routes workloads across 400+ models based on cost and performance. OpenRouter had 8 million users and raised $113 million at a $1.3 billion valuation just months ago, making the deal a major repricing of model-routing infrastructure. [TechCrunch]
  

• Amodei Says AI Backlash Is Really a Trust Problem ⚠️
  Anthropic CEO Dario Amodei pushed back on claims that AI leaders created today’s backlash, arguing that his warnings have consistently covered both risks and benefits. He sees the deeper issue as decades of declining trust in companies, governments, and technology, with AI now becoming the latest target of that broader distrust. [TechCrunch]

  

• Brockman Warns Companies to Move Fast on AI Security 🛡️
  OpenAI president Greg Brockman says the latest AI-driven security incidents show attackers will soon use agents to discover vulnerabilities at machine speed. His playbook calls for immediate assessments, AI-equipped security teams, faster forensics, and executive buy-in before the advantage shifts further toward attackers. [Business Insider]

  

• a16z Could Make Nearly $1.5B From OpenRouter Exit 🤝
  Andreessen Horowitz could turn roughly $20 million invested in OpenRouter into nearly $1.5 billion if Stripe closes its reported acquisition above $8 billion. Menlo Ventures could also clear more than $500 million, showing how quickly early-stage ownership can compound when infrastructure companies become strategic assets. [Business Insider]

• Cursor Launches Origin as GitHub Faces Outages 🖥️
  Cursor launched Origin as a collaborative code-hosting platform that syncs with GitHub, giving developers another layer for repositories, pull requests, and team workflows. The timing is notable: GitHub suffered a major outage the same day, highlighting the opportunity for developer platforms to reduce dependence on a single infrastructure provider. [TechCrunch]

• Sebastian Thrun Launches Robotics Startup Dulo 🤖
  Waymo pioneer Sebastian Thrun has unveiled Dulo, a stealth startup building foundation models for hardware design with talent from Waymo, Google Brain, and Stanford. The bet arrives as physical AI funding accelerates, with robotics startups raising $16.3 billion across 492 deals in Q1 2026 alone. [Business Insider]

• Flock Says Drones Are Its Fastest-Growing Business 🚁
  Flock Safety says drone response now accounts for nearly half of its forward-looking revenue, with more than 200 customers using drones as first responders to 911 calls. The growth is shifting the company’s center of gravity beyond license-plate readers, even as that core business faces scrutiny over accuracy, retention, and law-enforcement use. [Business Insider]

• Uber COO Predicts a Leaner Workforce From AI 🚗
  Uber COO Andrew Macdonald said the company could potentially operate its current business with fewer employees within five years as AI takes over more support, sales, content, and analytics work. The harder variable is new demand: AI may reduce headcount in existing functions while simultaneously making it economical for Uber to pursue work that previously required more people. [Business Insider]

• Anthropic’s IPO Is Being Priced on 2028 💰
  Anthropic is asking investors to underwrite roughly $190 billion to $200 billion in 2028 revenue, far beyond its $47 billion run rate disclosed in May. The valuation math is shifting from current earnings to future scale, with investors benchmarking the company against high-multiple names like Palantir, SpaceX, and Cloudflare. [Reuters]

New Funds 💰

• Reach Capital closed its fifth venture fund at $265M to back early-stage companies applying technology and AI across learning, health, and work.

• GD1 held a NZ$56.7M first close of its second venture fund to invest in Pre-Seed through Series B companies across deep tech, connected hardware, health tech, SaaS, and marketplaces.

• Third Mind is raising a $25M first venture capital fund to invest in early-stage technology companies.

• BlueHill Capital closed its inaugural venture fund at $41.8M to back Seed and Series A companies across energy, semiconductors, defence, space, robotics, and advanced materials.

• Odyssey Ventures secured a £10M commitment toward its $50M Discovery I venture fund, targeting seed and pre-seed AI and DeepTech companies.

That’s a wrap for this week.

If you want to stay ahead of the curve, make sure you're subscribed. No fluff, just real startup insights delivered straight to your inbox.

See you next time 🚀

This series is a preview of a new book I’ve written with my partners Richard Blundell and Paul Watson — The Selling Software Algorithm: The Go-to-Market Navigation System for B2B Software Leaders.

It isn’t a strategy deck dressed up as a book. Between the three of us we’ve spent close to seven decades inside B2B software as founders, operators, investors and coaches — building companies, exiting some, and quietly breaking a few along the way, which is where most of the real learning lives. Through our work at Vencha we’ve now supported hundreds of founders trying to do the single hardest thing in software: turn a clever product into predictable commercial traction in a market that’s noisy, cautious, and more crowded than it has ever been.

One pattern shows up again and again. Companies rarely struggle because the product is bad. They struggle because they have no navigation system for going to market. This book is that system.

Get my Copy

Most founders know their product is different. The problem is that their homepage often doesn’t make that difference obvious.

That matters more than it used to. Buyers are moving quickly, they have seen hundreds of versions of the same claims, and they are not going to spend five minutes working out what makes you different.

There is a very simple way to find out whether your message is doing its job. Take the hero message off your homepage and put it on your closest competitor’s. Then ask yourself a slightly uncomfortable question: would anyone notice?

That is where we’ll start.

Table of Contents

• A Simple Test: Swap Your Homepage With a Competitor’s

• Seven Seconds, and Not a Second More

• Why “Why” Beats “What” and “How”

• One Armour-Piercing Message, Not Five Soft Ones

• Strip Away the Algorithms and the Platforms

• The Modern Messaging Stack

• Matching Your Message to the Curve You Just Mapped

A Simple Test: Swap Your Homepage With a Competitor’s

Take the hero message off your website. Paste it onto your closest competitor’s homepage. Now ask your team whether anyone would notice.

If nobody would, the message isn’t yours. It’s category wallpaper, and it’s costing you every visitor who arrives without already knowing who you are.

The test works because it strips out everything you know about your own product and leaves only what the buyer can see. You know your architecture is different. They don’t, and they aren’t going to burn the mental calories finding out.

Seven Seconds, and Not a Second More

Buyers scroll the way everyone now scrolls. You have roughly seven seconds, often less, to produce the small internal jolt of recognition that says “that’s me.”

They don’t owe you patience. They don’t owe you curiosity. They certainly don’t owe you a meeting. Your messaging has to earn its place before it earns their attention, and most of it never does, because it was written to satisfy the founder’s pride rather than the buyer’s need.

There’s a deeper version of this too. Most software companies write messaging about operational pain: time saved, cost reduced, workflow streamlined. All true, all useful later as a convincer. But the great ones go one layer down, to the reputational risk, the internal politics, the recurring problem that keeps surfacing in leadership conversations. You’re not only selling operational improvement. You’re selling relief, credibility, and a better leadership story.

Why “Why” Beats “What” and “How”

“We are an AI-powered platform that uses machine learning to optimise X” isn’t a message. It’s a product description, and it forces the buyer to do the interpretation. Buyers no longer interpret. They discard.

Lead with “how” and you’ll lose even faster. “We unify data across systems using our proprietary engine” may be entirely true, and it’s still not the reason anyone will care.

Lead with why it matters, which usually means pain, consequence and outcome, in plain English, in the Economic Buyer’s own voice. If a twelve-year-old can’t read your homepage and explain what you do, the language is too complicated. Complexity isn’t sophistication. It’s usually fear, because vague language lets you hide and clear language forces you to commit.

From the book: what great hero messages sound like

One Armour-Piercing Message, Not Five Soft Ones

The most common failure here is the multi-message compromise. A company tries to speak to everyone, writes five value statements, dilutes each one slightly, and hopes the buyer will pick the one that fits.

Buyers don’t do that. They glance, they judge, and they leave.

You need one armour-piercing message, anchored in your number one pain. The second and third pains can sit underneath it, but the hero line leads with the sharpest blade you have: the knife in your product, the thing that changes their world most urgently.

If you’re unsure what that knife is, go back to your pain wheel rankings and ask a brutally simple question. Which pain, if removed, would make the Economic Buyer feel the greatest relief? That’s usually the knife.

Strip Away the Algorithms and the Platforms

There is a time to explain how you do it. It isn’t in the first seven seconds of someone reading your website.

Words like “platform”, “algorithm”, “proprietary engine”, “next generation” and “AI-powered” are, in early messaging, a form of hiding. They signal sophistication without stating meaning. Economic Buyers don’t buy sophistication. They buy business outcomes they can defend in a room full of people who weren’t on your demo.

Removing that decorative language isn’t dumbing down. It’s respecting their time.

The Modern Messaging Stack

Once your algorithm is clear, the structure that follows is straightforward, and it forces your language to stay buyer-first.

The hero message carries the top pain and the outcome. One line names who it’s for, narrow and specific, ideally naming the Economic Buyer outright. Three supporting proofs follow: a metric, a case study outcome, a time-to-value claim. Then a calm call to action, not “book a demo” but something that feels like the next logical step, such as “see how it works in your world”.

That sequence mirrors how buyers actually process information. Fast emotional recognition first, then credibility, then action.

From the book: the Modern Messaging Stack

Matching Your Message to the Curve You Just Mapped

Last time we mapped readiness, and the same discipline applies here. Not everyone inside your perfect customer profile is ready to hear the same thing.

Late adopters, particularly around AI, arrive nervous and unsure what to ask. They fear hallucinations, compliance exposure and reputational fallout, and they want proof, references and reassurance. Innovators and early adopters respond to clarity, urgency and an ambitious outcome, but they respond to honesty just as strongly. They want to know what’s possible, what’s proven, and what’s safe.

Your messaging should be built for the part of the curve you’re targeting first, which early on should almost always be that front section of your perfect customer profile.

This is exactly what Builder Step 5 exists to check: whether your hero message, your named buyer and your three proofs actually hold together, or whether you’ve written five soft messages and hoped one of them lands.

Selling Software Algorithm Builder, Step 5: Messaging (recreated illustrative mockup based on the tool’s documented output; a fictional example company is shown)

Next in the series, we look at what happens when the algorithm stops being a document and becomes behaviour, and why this is the one piece of work a founder cannot delegate.

-Chris Tottman

Most hero messages could sit on a competitor’s website without anyone noticing, because they were written to describe the product rather than change the buyer’s world.

The Selling Software Algorithm Builder Step 5 tests your messaging against your Economic Buyer and your three pains, and shows you exactly where it goes soft.

Start your 7-day free trial → Cancel anytime.

Test Your Messaging

Read more

Two founders walk into two different partner meetings this month. Both put up the exact same slide: 112% Net Revenue Retention. Same font, same green arrow, same confident pause after the number lands.

One of them keeps 94% of existing revenue and grows the rest through expansion. The other keeps 78% and covers the hole with aggressive upsell on two or three big accounts.

Same headline. One is a compounding business. The other is a leaky bucket that hasn’t sprung its last leak yet.

The partner in the room knows which is which within about ninety seconds, because they were never looking at the number on the slide. They were calculating the one you didn’t show them.

The gap between the two numbers is the signal. The number on the slide never was.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

The number hiding behind your number

The slide number is NRR, Net Revenue Retention: what you keep from existing customers, plus everything they grew into. Upgrades, seats, usage. Because expansion gets added back in, NRR is uncapped. You can print 130% with zero new logos.

The number the partner is calculating in their head is GRR, Gross Revenue Retention: the same base with expansion stripped back out. Churn and downgrades only. It can never go above 100%, which is exactly what makes it honest. It’s the floor of what you keep before anyone upsells anything.

Put both founders through the math:

▫️ Founder one: $4M ARR, loses $240K to churn, adds $720K in expansion. GRR 94%, NRR 112%. Gap: 18 points.

▫️ Founder two: $4M ARR, loses $880K to churn, adds $1.36M in expansion. GRR 78%, NRR 112%. Gap: 34 points.

A healthy gap sits between 15 and 25 points. Under 15 and you’re durable but leaving money on the table, which is a pricing problem, not a churn problem. Over 25 and you’re masking a leak with a handful of accounts that could all walk in the same quarter.

Most founders have never calculated GRR separately. They’ve only ever seen the blended number, and that’s rarely an accident. NRR is the number that goes up when you sell harder to the customers you already have, so it’s the one every dashboard defaults to and every board deck leads with. GRR just sits quietly underneath, refusing to be flattered.

What “good” looks like in 2026

Retention benchmarks are close to useless blended, because they move hard by segment. A 97% NRR is a red flag for an enterprise vendor and roughly average for an SMB one.

▫️ Enterprise (ACV $100K+): median NRR ~118%

▫️ Mid-market ($25K–$100K ACV): median NRR ~108%

▫️ SMB (under $25K ACV): median NRR ~97%

▫️ B2B blended: median NRR ~108%, GRR ~88%

▫️ Bootstrapped, $3M–$20M ARR: median NRR ~103%, GRR ~91%

▫️ Top quartile, any segment: NRR 125%+, GRR 94%+

This isn’t a vanity metric. Public SaaS companies above 120% NRR have traded at roughly 9.3x median EV/revenue, against 3.1x for those below 100%, according to Software Equity Group. Retention isn’t a line on your deck. It’s a multiplier on the entire company.

And it’s usually the first thing a partner sanity-checks against your growth story. A founder who shows up with 40% year-over-year growth and a soft GRR underneath it is telling the room, without meaning to, that the growth is bought rather than earned. The multiple follows the retention, not the other way around.

Your GRR is a fundraising number before it’s ever a retention number

A partner doesn’t ask about churn out of curiosity. It’s one input into a bigger judgment about whether you’re ready to raise at all, and it rarely travels alone. If retention is on your mind right now, these are worth reading next:

▫️ The Number That Kills More Fundraises Than Any Bad Idea: one slide, one metric, quietly killing raises without founders ever knowing it.

▫️ The Way VCs Actually Calculate Your Valuation: they already have a number in their head before you walk in. What is it, and how do you hit it?

▫️ How Investors Decide If You’re Ready to Raise: it’s not timing, it’s not luck. The exact criteria investors use.

▫️ The Quiet Filter That Decides Your Entire Fundraise: every investor runs this before you open your mouth.

Which of the four quadrants are you in

Plot yourself. GRR on one axis, the gap on the other. There are only four places to land, and each one gets a different verdict in a partner meeting.

▫️ GRR 90%+, gap under 15 points, durable and under-monetised. People stay, you’re just not capturing the value yet. Fastest fix here: move to usage- or seat-based pricing so revenue climbs alongside the customer’s own success.

▫️ GRR 90%+, gap 15 to 25 points, the compounding machine. Broad retention, broad expansion. This is the profile that earns the premium multiple, and the only quadrant where churn genuinely stops being a worry.

▫️ GRR under 85%, gap 25+ points, the masked leak. The headline looks fine. The base is bleeding and a handful of accounts are carrying the whole story. Partners are trained to find this quadrant, and them finding it is far worse for you than you disclosing it first.

▫️ GRR under 85%, gap under 15 points, honest and broken. No expansion mechanic, real churn. Painful, but the cleanest diagnosis on the board: you know exactly what to fix, and nobody can accuse you of hiding it.

Most founders assume they’re in quadrant two and have simply never run the GRR math to check.

The fix you can start this week, before touching the product

Here’s the part almost nobody acts on, because it doesn’t feel like a growth initiative.

Across analyses from ProfitWell, Recurly and Paddle, failed payments account for 20 to 40% of all SaaS churn. Expired cards. Bank declines. Customers who never chose to leave and, in a lot of cases, never even noticed they’d gone. A proper dunning setup recovers 70 to 80% of them, per Chargebee’s research on the subject.

This is the highest-leverage fix on the list, because it’s a technical problem with nobody’s opinion attached. You’re not persuading anyone to stay. You’re just catching the ones who wanted to.

▫️ Retry on a schedule, not once. Days 1, 3, 5 and 10 beats a single attempt at the moment of failure, since most declines are timing issues that clear within a week.

▫️ Email before the card expires, not after. Pre-dunning recovers far more cheaply than chasing a failure after the fact.

▫️ Turn on the account updater. Stripe and Adyen will refresh reissued card details automatically. Most teams have simply never flipped the switch.

▫️ Put it in the app, not just the inbox. Billing emails land in a finance address nobody checks. The person who can fix it is the one logged in.

▫️ Split the number in your own reporting. Report voluntary and involuntary churn separately from here on. If you can’t split them yet, that’s the tell you don’t know which problem you actually have.

▫️ Report GRR trailing twelve months, not last month. A single bad month reads as a crisis. Twelve months reads as a trend, which is the only version a partner actually trusts.

One to two weeks of engineering. Zero new customers acquired. Several points back on the number that was carrying the real signal the whole time.

🔒 This is exactly the kind of asset we built Premium for

You now have the diagnostic and the fastest fix. What you don’t have yet is the other side of the table, and that’s the part that actually decides your round.

Behind the paywall is The Retention Diligence Pack. The nine questions a partner asks about your retention, in the exact order they ask them, with the answer that lands and the answer that quietly ends the process, including the one about your largest customer that catches out almost every founder who hasn’t rehearsed it. The cohort table format built for a data room, the one that makes a partner stop probing instead of asking twenty more questions. And the three ways founders accidentally disclose a churn problem they don’t actually have, one of which is a single word in an investor update that has cost people term sheets.

Plus the NRR/GRR gap calculator itself, pre-built so you can drop in your own numbers and know which quadrant you’re actually in before anyone else tells you.

Premium subscribers get:

✅ The Retention Diligence Pack: nine diligence questions with model answers, the cohort table template, and the three silent disclosure mistakes

✅ The NRR/GRR gap calculator, ready to drop your own numbers into

✅ Every past and future proprietary framework, model and swipe file we publish

✅ Early access to new research before it goes out to the free list

Paid subscribers also get:

✅ 60+ additional tools across every stage of fundraising and company building: investor research, pitch deck screening, meeting prep, term sheet analysis, financial models, and the full investor database library.

Start your 7-day free trial → Cancel anytime.

Unlock the Full Retention Diligence Pack

Read more

One number in it should change how you run your raise.

87.5% of all US venture dollars in the quarter went to AI companies. Every other sector in technology, fintech, healthtech, climate, marketplaces, consumer, devtools, logistics, split the remaining 12.5% between them.

It is the most skewed AI-versus-everything split PitchBook has ever recorded.

The single week the report covers makes it concrete. Lovable raised $400 million at a $13.3 billion valuation. River AI raised $1.1 billion across seed and Series A, two months after leaving stealth. CodeRabbit closed $143 million at $1.5 billion. Cognition opened talks on a round that could value it above $40 billion.

Four companies. One week. All AI.

Zoom out and it gets stranger. Global venture funding hit a record $510 billion in the first half of 2026, more than all of 2025 combined. OpenAI and Anthropic alone took more than 40% of it. In Q1, roughly 65% of every venture dollar on earth went to four companies.

If you are not building AI, you have spent the last eighteen months being told, in a hundred small ways, that the money left the room.

It didn’t. But where it went is not where you have been looking.

Want to get in front of 250k+ founders and investors?

For sponsorship opportunities across The Founders Corner newsletter and LinkedIn (160k followers), email: founderscornermedia@gmail.com

The number nobody did the math on

Percentages are designed to make you feel something. Absolute numbers are designed to tell you something.

Crunchbase puts total Q2 2026 investment into US and Canadian startups at $137.2 billion.

Run 12.5% against that and you get roughly $17 billion. Deployed into non-AI companies. In ninety days.

For scale: that is more capital than the entire US venture market deployed in some full years of the early 2010s. It is not a rounding error. It is a functioning market that happens to be standing next to a much louder one.

(The two datasets do not measure identically, so treat $17 billion as an order of magnitude rather than a decimal point. The direction is not in dispute.)

Here is what actually changed, and it is not “there is no money for you.”

The money didn’t shrink. The default search path broke.

The generalist multi-stage funds that used to be the obvious first fifty names on a non-AI target list are now under enormous LP pressure to show AI exposure. They will still take your meeting. They are far less likely to lead your round. Founders keep pitching that list, keep getting warm passes, and conclude the market is closed.

The $17 billion did not come from that list. It came from nine other places.

If you’re raising right now, you’ll need to read these:

• The Number That Kills More Fundraises Than Any Bad Idea: one slide, one metric, quietly killing raises before founders even notice.

• The Quiet Filter That Decides Your Entire Fundraise: every investor runs this before you open your mouth.

• The Way VCs Actually Calculate Your Valuation: they have a number in their head before you walk in.

• How Investors Decide If You’re Ready to Raise: it’s not timing, it’s not luck, it’s a checklist.

Which of the three buckets you are actually in

Before you touch your target list, be honest about which market you are raising in. There are three positions and only two of them are viable.

Bucket 1: genuinely AI-native. The model is the product. Remove AI and there is no company. You are competing in the 87.5% market, and the bar there is brutal but the capital is limitless.

Bucket 2: honestly non-AI. You use AI internally, like everyone does, but you sell software, hardware, or a service that would exist without it. You are competing in the 12.5% market, where there is $17 billion a quarter and considerably less noise.

Bucket 3: AI-washed. AI is in the deck. It is not really in the product.

Bucket 3 is the one that kills raises, and it is the most crowded bucket in the market right now.

The logic is simple and most founders get it backwards. By putting AI at the front of your narrative, you volunteer for the 87.5% market. Once you are in it, you get measured against companies where the model is the product. “We use GPT” raised money in 2023 and raises nothing in 2026. Investors now ask one question in the first meeting: what happens when OpenAI ships this? If your AI is a feature, you have no answer, and you lose a race you never needed to enter.

Meanwhile the 12.5% market, where you would have been a credible category leader with real margins and a defensible customer relationship, never got to see you.

Being honestly non-AI is a stronger position than being unconvincingly AI. That sentence is worth more than most fundraising advice published this year.

The 9 investor categories still writing non-AI checks

These are the nine places the $17 billion actually came from. Two of them, in full, below.

Category 1: Vertical sector specialists

Funds whose entire thesis is a non-AI sector: industrials, insurance, construction, food systems, logistics, marine, energy retail. Their LPs did not give them money to chase foundation models, and they are quietly delighted that the generalists have vacated their patch.

What they screen for: domain credibility over technology novelty. They want to know you have sold into this industry before, or that someone on your cap table has. They will forgive a slower growth curve for a customer relationship they understand.

How to reach them: they are almost never on the standard “top VCs” lists your peers are working through. They show up as sponsors at industry conferences, not tech conferences. That is the search term.

Category 2: Family offices

Single-family and multi-family offices are structurally the best-positioned buyer in this market. No LP to answer to, no fund cycle forcing deployment into whatever is hot, and often an operating business in exactly the sector you are selling into.

What they screen for: durability and downside. They ask what the business looks like if it never raises again. Answer that credibly and the conversation moves faster than any VC process you have run.

How to reach them: they do not have a submission form and they do not want one. Access runs through the operating business, the wealth manager, or a portfolio founder. Warm-intro discipline matters more here than anywhere else on this list.

Categories 3 through 9, including the two that write the largest non-AI checks in the market right now and the one most founders have never seriously considered, are below.

🔒 You now know the market exists. Here is the map to it.

Everything above tells you the $17 billion is real and which bucket you are raising in. What follows is the part you can put to work tomorrow morning. Here is exactly what is waiting below:

✅ The full 9-category investor map, as a single reference table: who they are, the check size band they actually write, what each one screens for, and the specific route in. Categories 3 to 9 include corporate venture arms, revenue-based and non-dilutive capital, regional and sovereign-backed funds, lower-mid-market growth equity, operator holdcos, Fund I and Fund II emerging managers, and strategic customer-funded development.

✅ The one line that decides which market you are raising in. There is a single sentence near the top of your deck and the top of your outreach email that sorts you into the 87.5% market or the 12.5% market. Most founders write it wrong and never find out why the meetings went cold. The corrected four-part frame, with a worked before-and-after example.

✅ Three copy-paste Claude prompts: one that audits your existing deck for accidental AI-washing, one that rebuilds your positioning line against the frame, and one that generates a ranked non-AI investor target list from your actual sector and stage.

✅ The capital sequencing rule for a 12.5% raise, which is a different order of operations to a standard venture process and the reason most non-AI raises stall at week six.

Paid subscribers also get:

✅ 60+ additional tools across every stage of fundraising and company building: investor research, pitch deck screening, meeting prep, term sheet analysis, financial models, and the full investor database library.

Start your 7-day free trial → Cancel anytime.

The one line that decides which market you are raising in

Open your deck. Find the sentence that describes what your company is. It is usually on slide one or slide three, and it is the same sentence you use in the first line of every cold email.

That sentence is doing the sorting.

The 9-category map

Read more