Why your seed deck is three slides too long

You have a draft. Twenty-something slides, the content mostly in place, and you want to know if it's time to make it look good. My answer: first, take three slides out.

Three is a guess. Your number might be two, or five. What I'm fairly sure of is that the draft contains slides that exist to make you feel covered, and that those slides sit between an investor and the answers they came for. Finding them is a narrative job. It belongs before design, and the sooner you do it, the cheaper it is.

What an investor actually does with your deck

Start with the reader. The figure most often quoted comes from a 2015 study DocSend ran with Harvard Business School's Tom Eisenmann: investors spent an average of 3 minutes and 44 seconds on a deck. DocSend's later reports put it lower, under three minutes in 2022. The exact number moves from year to year. The shape doesn't. It's a few minutes, and it's a skim.

Even the slide that held investors longest in that study, the financials, got about 23 seconds. DocSend's later seed reports describe investors zeroing in on market fit, monetization and momentum. Put that together and you get a picture of someone skimming toward a few specific answers. Every slide between the cover and those answers is a toll they pay with attention. Some of them won't pay it.

The decks in the 2015 study averaged 19 slides. That describes what was common. It isn't a goal, and hitting it proves nothing about whether the story works.

The three slides I would cut first

The extra slides tend to be the same ones. Here they are side by side.

Slide	Why it's there	What it costs the reader	What replaces it
Industry history lesson	Context, comfort, relevance	Delays the pain point	One concrete customer failure
Feature roadmap	Proof of ambition and technical depth	Blurs what the product is	The single mechanism that works today
Multi-stream revenue model	A bigger-looking market	Hides how the first dollar arrives	One model, one market, tested numbers

The history lesson explains how the industry works, what's broken in general, which macro trends are moving. An investor who covers your sector has heard this explained many times. Your version delays the one thing they can't get from anyone else: the specific, urgent failure your customer lives with. Replace the lecture with that failure, described concretely enough that the reader can picture who suffers it.

The roadmap slide lists modules, sub-features and a three-to-five-year plan nobody has validated. Founders include it to show they think big. At seed, the question in the room is narrower. Does one mechanism solve one beachhead problem? Five modules make that question harder to answer, because the reader can no longer say what the product is in a sentence.

The revenue slide with five streams (subscription, marketplace take, enterprise licensing, data, something else) reads to the founder like a larger market. To the investor it reads like an open question: how do you make money first? Show one business model for the first market, with unit economics grounded in what you've actually tested. The other streams can wait for the question.

All three answer a worry the founder has. None of them answers a question the investor is asking.

Why the cut comes before design

My process, which I call From Story to Investment, runs in three phases. The first is teardown and assembly: who is in the room, what a yes looks like for this particular meeting, and an argument map of what the deck says, what it leaves out and in what order. The second is narrative building: a written narrative document, the deck structure, and a core message for each slide. Design and delivery come third. For a fundraising pitch the whole thing usually takes about two weeks.

The cuts happen in phase one, for a practical reason. Designed slides are harder to kill. They look finished, and finished things feel true. Once someone has spent an afternoon on a market diagram, the conversation shifts from whether the slide belongs to how to keep it.

AI tools sharpen the problem. They make it cheap to produce a polished slide, which is useful. They don't make it cheaper to decide what the deck should say, to whom and in what order. A fast tool can even make it easier to keep the wrong slides, because each one arrives already looking good.

Settle the story in plain text. Then design it.

Fewer ideas, sometimes more slides

Cutting is about ideas. Page count follows, or occasionally doesn't.

The Neuromagen Pharma investor pitch shows how. It was a 15-minute presentation to about 150 pharma investors on a two-day New York program where fifty companies presented. In a room like that, a market explainer is the slide most likely to repeat what the audience already knows. We cut the general market and drug-development background. The opening came from something the CMO mentioned in passing: the founder had spent 25 years on a single compound. That fact did more work than any background slide could.

We also split overloaded slides so each carried one idea. Splitting adds pages, and that's fine. A slide with one idea is faster to take in than a slide with three, even if there are more of them. Afterwards, conversations opened with several investors. The deck didn't make that decision for anyone. It made the company easier to understand in fifteen minutes, which is the part a deck can actually do.

Where the cut slides go

Cut doesn't always mean deleted. The roadmap and the extra revenue streams often belong in an appendix, ready for the question when it comes. If an investor asks about your three-year product plan, you want the answer one slide away. You just don't want them to read it before they know what you do today.

How hard you cut also depends on which deck you're building. A presenter deck supports you while you talk, so it can lean on your voice and carry less text. A send-ahead deck has to carry itself on a screen with nobody there to skip ahead for the reader. Every slide needs a headline that states a full claim, and every extra slide is one more place for a solo reader to stop.

The WindowFix send-ahead partnership deck was built that way: fifteen slides for a Swiss architectural glazing manufacturer, opened as a conversation between equals instead of a company profile, and narrowed to private luxury villas so the message didn't scatter. Narrowing is cutting too. It removes the parts of the market you aren't talking to.

How to find your three

You can do this yourself, in a document, before anyone opens a design file.

1. Write down what a yes looks like for this meeting, in one sentence. A second conversation, a partner meeting, a data room request. Be specific.
2. Rewrite every slide title as a full-sentence claim. If you can't say what a slide claims, it's a candidate for cutting.
3. For each slide, ask whether an investor in your sector needs it to reach that yes. If they would already know it, cut it.
4. Look at everything between your cover and the slides that show your market, your business model and your momentum. Anything that doesn't move the reader toward those answers is in the way.
5. Read the headlines alone, in order. If the story holds without a slide, the slide goes, or moves to the appendix.

The last step is where founders get honest with themselves. The slides you defend hardest deserve a second look, because some of them are protecting you from a question you'd rather answer live.

If you've done all this and you're stuck between two slides, a second reader helps. I offer a free 30-minute call to look at a deck in progress, and you can see how a settled narrative turns into finished pages in slides from decks I've designed.

Three slides is a guess, and your number may differ. The habit is what matters: decide what the people in the room need to believe, remove whatever doesn't help them get there, and only then start designing. A shorter story is easier to follow, and an investor who can follow it reaches their own decision sooner, whichever way it goes.