Lesson 8 of 8

Now Write the Pitch (With Evidence Behind Every Claim)

The pitch is the last thing you write, not the first. Assemble your answers from lessons 1-7 into a short narrative where every sentence has a receipt.

8 min read

There is no new investor question in this lesson, because you have already answered the ones that matter. Is the problem real. Who has it. How big is it. Will they pay. Why you over the alternatives. How you reach them. Most founders write the pitch first and go hunting for evidence to decorate it. You did it in the right order, which means the pitch is no longer a writing problem. It is an assembly problem, and this lesson is the assembly manual.

The pitch comes last for a reason

A pitch written before the evidence exists is fiction with formatting. That is what most seed decks are: eighteen slides of adjectives arranged around a hole where the proof should be. Investors read hundreds of these, and they have learned to skim past the vision language looking for one thing: has this founder actually touched the market, or only imagined it.

You are in a different position now. If you worked lessons 1 through 7, you have seven evidence-backed answers sitting in your notes: a problem people described in their own words, a bottom-up market number, demand you measured, money that moved, a named buyer, a wedge against real alternatives, and a channel you tested. The pitch is those seven answers, in order, with an ask at the end. Nothing else needs to be invented.

This also changes how the pitch feels to deliver. Founders who are reciting hopes get defensive under questioning, because every question threatens the story. Founders who are reporting results get curious under questioning, because every question has an answer they already found. Investors can tell the difference in the first two minutes.

Map your answers onto the standard narrative

Investor pitch structure is boringly standard, and that is good news: you do not need a creative format, you need to slot your evidence into the one they already expect. Problem, who has it, market, proof of demand, willingness to pay, competition and your wedge, go-to-market, the ask. Each beat maps to a lesson you have already done the work for.

The discipline is one or two sentences per beat, each carrying its strongest piece of evidence. Not the whole study. Not every quote. The single number or behavior that would make a skeptic pause, with the rest held in reserve for questions. A short pitch full of receipts beats a long pitch full of adjectives, every time.

Problem and who has it: your interview evidence (lesson 1) plus your named ICP (lesson 5). One sentence of pain, one sentence of buyer.
Market: your bottom-up number and how you built it (lesson 2). Never the top-down trillion.
Proof of demand and willingness to pay: what people did, then what they paid or committed (lessons 3 and 4). This is the spine of the pitch.
Competition and wedge: what your buyers use today and the specific reason they would switch (lesson 6).
Go-to-market and the ask: the channel you tested (lesson 7), then what the money buys and which gaps it closes.

The honesty audit

Before the pitch leaves your laptop, go through it claim by claim and mark each one E or H: evidence-backed or hope. E means you can point to a thing that happened, an interview, a payment, a conversion rate, a measured cost. H means you believe it, and that is all. Be brutal, because the alternative is worse: investors run exactly this audit in the first five minutes of diligence, and they are better at it than you are.

For every H, you have two honest moves. Close it: if the test is cheap and takes a week, run it before you pitch, the same way you ran the tests earlier in this course. Or state it openly: name the gap in the pitch and make it part of the ask. We have not proven retention past month three; the raise funds twelve months of cohort data. That sentence does more for your credibility than any adjective, because it shows you know where your own story is thin.

What you must not do is leave an H dressed up as an E. One inflated claim, discovered in diligence, poisons every real number in the deck. The whole value of the work you have done is that your pitch can survive a skeptic. Do not spend that on a sentence you could have just tested.

Say it in two minutes

Write the spoken version first: a two-minute narrative, roughly ten sentences, one beat each, every sentence traceable to evidence. The deck comes after, one slide per beat, and it will be short because you are not padding around missing proof. If you cannot say it in two minutes, that is usually a sign a beat is missing its evidence and you are compensating with words.

End with the ask, and tie it to the audit: this is what we have proven, these are the two things we have not, and this raise exists to prove them. That framing turns your gaps from weaknesses into the literal purpose of the round, which is what a pre-seed or seed raise actually is.

If you ran your idea through an Olune validation run along the way, the report it produces is effectively the middle of this pitch: the demand evidence, the competitor read, and the honest verdict, in one place. Either way, the standard stays the same: no sentence goes in the pitch without a receipt behind it.

Draft the two-minute spoken version before you open a slide tool.
One beat per sentence or two; hold the detail for questions.
Mark every claim E or H, then close or confess every H.
Make the ask specific: which unproven claims does this money prove, over what runway?
Rehearse the questions, not just the monologue. Your evidence is your answer bank.

Worked example

Maya's final two-minute pitch, with receipts

Maya's first deck, written before any of this work, was sixteen slides that opened with the future of agency operations. She deleted it and wrote ten sentences instead, each one pulled from evidence she gathered in the earlier lessons. It sounds like this.

Small marketing agencies burn most of a working day onboarding each new client: chasing signatures, briefs, and access credentials across email and spreadsheets. I interviewed 24 agency owners and 19 described this same manual mess unprompted; 11 showed me the duct-taped workarounds they built themselves. My buyer is the 5-to-20-person agency signing several new clients a month, and my bottom-up math puts the reachable English-speaking segment in the low tens of thousands of agencies. I tested demand with a landing page and paid pilots: cold traffic converted to a waitlist at 11 percent, three agencies prepaid 500 each for a pilot, and six in total now pay 99 a month for a version I largely run by hand. They came off generic e-signature and project management tools, which know nothing about agency onboarding; my wedge is workflows built for exactly that. My first channel is the two agency communities my pilot customers came from, where I have measured what it costs me to start a sales conversation. I am raising enough for eighteen months to prove the two things I have not: retention past the early months, and a second acquisition channel.

Every sentence traces to something that happened: interviews, a landing page test, paying pilots, a competitor teardown, a channel experiment. When an investor pushes on any line, Maya does not defend, she elaborates, because there is material underneath. That is the entire difference between this pitch and the sixteen slides she deleted.

Learn by doing

Paste these into ChatGPT or Claude and run them against your own idea. The model will answer happily. Olune goes further and checks the answer against real Reddit threads, competitor maps, and keyword volume.

Prompt 1 · Assemble your evidence into a two-minute pitch narrative.

I am writing an investor pitch for [YOUR IDEA IN ONE SENTENCE]. Here are my evidence-backed answers: problem evidence: [WHAT YOUR INTERVIEWS SHOWED], buyer: [YOUR ICP], market: [YOUR BOTTOM-UP NUMBER AND HOW YOU BUILT IT], demand proof: [WHAT PEOPLE DID], willingness to pay: [WHAT PEOPLE PAID OR COMMITTED], competition and wedge: [WHAT THEY USE TODAY AND WHY THEY WOULD SWITCH], go-to-market: [THE CHANNEL YOU TESTED AND WHAT IT COST]. Write a two-minute spoken pitch of roughly ten sentences, one narrative beat per sentence, using only the evidence I gave you. Do not add claims I did not supply. End with an ask that names what remains unproven.

What a good output looks like

For Clientfile it returns a tight narrative: the onboarding pain in the agencies' own words, 19 of 24 interviews confirming it, the 5-to-20-person agency buyer, the bottom-up market figure, six paying concierge pilots at 99 a month as the demand spine, the wedge against generic e-signature tools, the community channel with a measured cost per conversation, and an ask framed around proving retention and a second channel. Anything Maya left blank, the draft flags as missing evidence instead of papering over it.

Prompt 2 · Run the honesty audit on your draft pitch.

Here is my draft investor pitch: [PASTE YOUR PITCH]. Act as a skeptical investor doing diligence. Go through it claim by claim. Mark each claim E if I supplied concrete evidence for it (something that happened: interviews, payments, conversion numbers, measured costs) or H if it is hope dressed as fact. For every H, tell me the cheapest test that would convert it to an E within two weeks, or how to restate it honestly as a known gap the raise will close.

What a good output looks like

On Maya's draft it marks the interview findings, the paying pilots, and the channel cost as E, then flags two H claims: 'agencies will stay past onboarding season' (fix: restate as a named gap the raise funds, since the cohort data needs months) and 'we can expand into design studios' (fix: cut it, or run five interviews with studio owners before the pitch). It also catches one inflated verb: 'dozens of agencies use Clientfile' becomes 'six agencies pay for the concierge version', which is smaller and far more convincing.

Key terms in this lesson

How to Write a Business Plan for Investors How to Write a Business Plan (The Lean Version)

Takeaways

Write the pitch last. It is an assembly of your evidence, not a work of imagination.
One beat per lesson: problem, buyer, market, demand, money, wedge, channel, ask.
Audit every claim as evidence or hope. Investors will run the same audit in minutes.
Close the hope claims with cheap tests, or name them openly as what the raise is for.
A short pitch full of receipts beats a long pitch full of adjectives.

Now run your own idea through it.

You have the method. Olune does the legwork: an honest build-or-kill verdict on live Reddit signals, competitor maps, and keyword volume, in about eight minutes. Free to start.

Test your idea →

Common questions

How long should the pitch actually be?

Two minutes spoken, and a deck of roughly ten slides, one per narrative beat. Length in a pitch is almost always compensation: founders pad where the evidence is thin. If yours runs long, find the beat that is hiding a hope claim and fix that instead of the wording.

What if my honesty audit comes back mostly H?

Then you are not ready to pitch, and finding that out now is the win. Go back to the lessons whose evidence is missing and run the cheap tests first; most of them take a week or two. A month of evidence-gathering will do more for your raise than a year of deck polish.

Will admitting gaps make investors pass?

Some, and those were going to find the gaps anyway. A named gap with a plan reads as self-awareness; a hidden gap discovered in diligence reads as either sloppiness or dishonesty, and both end the conversation. Early-stage investors expect unproven claims. What they are pricing is whether you know which of your claims those are.