The Seed Deck and Narrative Playbook

A deck is the visual form of an argument. Founders who open PowerPoint first end up with twelve well-designed slides that do not add up to a reason to invest. Write the argument as four paragraphs of prose before you make a single slide. If the prose is unconvincing, better design will not fix it.

The argument at seed has four load-bearing claims.

1. This problem is real and expensive, and here is evidence beyond your own conviction.
2. The way it gets solved is changing, because of something that became true recently.
3. We are the ones who solve it, for a reason that is hard to copy.
4. Early evidence says it is working.

Everything else on a slide is supporting material. If a slide does not advance one of these four claims, cut it.

The "why now" is the seed-stage crux

At Series A, traction carries the round. At seed, you frequently do not have enough traction to be dispositive, so the burden shifts to the insight.

Investors are asking: why is this possible or necessary now, when it was not three years ago? Valid answers:

• A technology capability crossed a threshold (cost, accuracy, latency, availability)
• A regulation changed or is about to
• A behavior shifted at scale
• An incumbent's business model now prevents them from responding
• A cost curve inverted

Invalid answers: the market is growing, AI is big, more people are online now. These are true of everything and therefore say nothing.

The strongest "why now" is one the investor had not thought of and cannot argue with once you say it. This is the single highest-leverage sentence in your entire pitch. Spend disproportionate time on it.

The slide sequence

Ten to twelve slides. Each one readable in fifteen seconds without you talking, because it will be forwarded and read without you.

Problem

Name a specific person having a specific bad day. Generic market pain reads as desk research. If you lived the problem, that is your credibility and it belongs here.

Solution

Show the product. A screenshot beats a paragraph of description every time. Founders consistently under-show the actual thing they built.

Traction

Any chart with time on the horizontal axis. If the number is small, show it anyway with the growth rate. Small and compounding beats vague and impressive. If you have no revenue, show the strongest proxy: usage, retention cohorts, waitlist conversion, signed pilots, letters of intent.

Market

Build it from the bottom: number of qualifying customers, multiplied by realistic annual contract value. A $50B TAM from an analyst report is treated as noise. A defensible $2B built from named customer segments is treated as thinking.

Competition

Never claim you have none. It reads as either naive or dishonest, and the investor will name three the moment you leave. Include the status quo (spreadsheets, an internal script, doing nothing) as a competitor, because it usually is the real one. A two-axis positioning chart is acceptable if the axes are the two things a buyer actually decides on.

Team

Answer "why you" rather than listing resumes. Prior work at a recognizable company is one line, not three. Founder-market fit, meaning why this specific team has an unfair understanding of this specific problem, is what the slide is for.

The ask

Amount, instrument, and what it buys. Not a generic pie chart of spending categories. State the milestone: "This gets us to $1.5M ARR and 20 enterprise logos, which is a Series A." Know whether you are raising on a SAFE or a priced round before this slide exists.

Two decks, one story

The send deck stands alone. More text, self-explanatory, designed to survive being forwarded to a partner who never met you. This is the one you email.

The present deck is sparser. Fewer words, bigger visuals, because you are the narration. This is the one on screen when you are in the room.

Same argument, different density. Maintaining both is worth the effort because the failure modes are opposite: a send deck with six words per slide is useless forwarded, and a present deck with paragraphs makes you read your own slides aloud.

The verbal versions

You need the same argument at four lengths. Practice each until it is automatic.

Ten seconds. "We do [X] for [Y]." No adjectives, no vision, no category creation. Someone should be able to repeat it correctly to a colleague.

Sixty seconds. Problem, why now, what you built, one traction number, what you are raising.

Five minutes. The four-claim argument, with one piece of evidence per claim.

Twenty minutes. The full walkthrough, expecting interruption. Build it so it survives being taken out of order, because it will be.

The objection playbook is what you use when the twenty-minute version gets interrupted, which it will.

Testing the narrative

Before you take real meetings, run the pitch past three audiences and listen for different things.

• Another founder in your space. Are you wrong about anything factual?
• An operator who is your target buyer. Is the problem framed the way they would frame it?
• Someone with no context. Where did they get lost? First confusion point is your weakest slide.

Then run five practice meetings with investors you do not want money from. Track which question you answer worst, fix it, repeat. Never make your best target your first meeting. How to Run a Seed Round covers how those meetings sit inside the rest of the process.

What the deck is not

• Not a business plan. Nobody reads the 40-page version.
• Not a product tour. Five product screenshots is four too many for slide 5.
• Not a design showcase. Clean and legible beats beautiful. Bad design costs you nothing at seed. Unclear thinking costs you the round.
• Not a substitute for the conversation. Investors decide on the person and the insight. The deck earns the meeting and survives the forward.

Failure modes

1. No "why now." The most common fatal gap in seed decks.
2. Top-down TAM. Signals you have not thought about who actually buys.
3. Claiming no competition. Reads as unserious.
4. Hiding the product. If there is no screenshot, investors assume it does not exist.
5. Vision before evidence. Earn the ten-year story with the twelve-month numbers.
6. Too many slides. Twenty-five slides means you could not decide what mattered.
7. A deck that needs you present. It will be read without you, by the person who decides.