How to Make a Good Pitch Deck: The 12 Slides Investors Actually Read

A pitch deck has one job: earn the next meeting. It does not close the round, explain every feature, or answer every objection — it has a few minutes to convince a busy investor that your team, your problem and your early proof are worth an hour of their time. This guide walks through the twelve slides that do that job, in order — team, problem, solution, product video and the eight that follow — with what each slide must answer, what to leave out, and the mistakes that get decks closed early.

The most useful number in fundraising comes from a 2015 study by DocSend and Harvard Business School professor Tom Eisenmann, which tracked how investors actually read the decks of 200 startups that together raised about $360 million. Investors spent an average of 3 minutes and 44 seconds on a deck. The average deck ran to 19 pages, founders needed around 40 investor meetings, and a round took a little over 12 weeks to close, as TechCrunch reported at the time.

Read those numbers together and the design brief writes itself. Nobody studies your deck; they skim it, usually alone, often on a phone, between two other things. Every slide gets a few seconds to make one point. If the point is not obvious from the headline, it is not made.

It also means you need two versions of the same deck. The send-ahead version is read without you in the room, so each slide must stand alone: a full-sentence headline and just enough text to make sense cold. The presentation version is the backdrop while you talk, so it carries far less text — you are the narration. Founders who send the presentation version by email are sending a deck that makes no sense without them.

There is no single official order. Sequoia Capital's long-standing template runs company purpose, problem, solution, why now, market, competition, business model, team, financials, vision. Y Combinator's seed deck guide is shorter still. The order below puts the team first, which is a deliberate choice for early-stage companies: before there is revenue, the people are the strongest evidence an investor has. The same DocSend study found that team, financials and competition were the slides investors spent the longest on. If you already have exceptional traction, move that slide up to the front instead — lead with whatever your strongest proof is.

The question it answers: why are you the people who will win this? Not “who are you” — investors can find your LinkedIn. What they cannot find is the connection between your background and this specific problem. That connection is often called founder–market fit, and this slide exists to prove it.

• Show two to four people: founders and genuinely key hires. Photo, name, role, and one line each.
• Make that one line an achievement relevant to this company — “ran payments infrastructure for 4M users” beats “10+ years of experience”.
• Show that the team covers the essentials between them: someone who can build it and someone who can sell it.
• Say who is full-time. Investors will ask, and a part-time founding team is a fact better disclosed than discovered.
• Leave advisors off unless they are actively involved and their name genuinely changes the conversation.

The question it answers: who is in pain, and how much does it cost them? A strong problem slide names a specific person, describes what they do today to cope, and puts a number on what that workaround costs in money, hours or lost customers. Sequoia's guidance is to describe the pain of the customer and how they address the issue today — because the workaround is the proof that the pain is real. People do not build spreadsheets and hire assistants for problems they do not have.

Keep it to one problem. Three problems on one slide reads as a team that has not yet decided what it is building. And use the customer's own words if you have them: a single quote from a real interview is more persuasive than a paragraph of your own description.

The question it answers: what do you do, in one sentence a stranger would understand? State the outcome for the customer, not the technology underneath. “Landlords collect rent and chase late payers automatically” is a solution. “An AI-powered, cloud-native property platform” is a category label that could describe fifty companies.

The best solution slides mirror the problem slide directly: the cost you quantified one slide earlier is the cost you now remove. Resist the feature list. Three benefits at most, each tied back to the pain, and then let the next slide do the showing.

The question it answers: is this real? Nothing in a deck builds belief as quickly as watching the product work. A short demo video moves you from “claims on slides” to “software that exists”, and for a pre-revenue company that shift is often worth more than any other slide.

• Keep it to 60–90 seconds. One user, one job, start to finish. It is a demonstration, not a tour of every screen.
• Be inside the product within five seconds. No logo animation, no founder introduction, no stock footage of people shaking hands.
• Show the real product. If parts are a prototype or a mockup, say so. Investors forgive early; they do not forgive discovering later that the demo was staged.
• Narrate the outcome, not the clicks. “The tenant has now paid and the landlord's books are updated” rather than “now I click Submit”.
• Add captions and assume it may be watched with the sound off.
• Remember that a PDF cannot play video. Put a clean product screenshot on the slide with a visible play button, and link it to an unlisted video that opens without a login. Test the link from a phone that is not signed in to anything.
• For a live pitch, keep the file on your laptop. Never let the most convincing minute of your pitch depend on a meeting room's Wi-Fi.

No product yet? Use a clickable prototype recording and label it as one, or replace this slide with two or three annotated screenshots. What you should not do is skip the showing entirely — a deck with no product visuals asks the investor to imagine the product for you, and they will imagine something less good than what you have.

The question it answers: why has nobody already done this, and why will it work today? Good ideas are usually old ideas whose timing has finally arrived. Name the specific change that opens the window: a new regulation, a cost curve that crossed a threshold, a shift in behaviour, a platform that did not exist three years ago. Sequoia treats this as its own slide for a reason — if there is no answer to “why now”, the investor's default assumption is that the idea has been tried and has failed.

The question it answers: if this works, how big can it get? Present the market in the three conventional layers — TAM (everyone who could conceivably buy), SAM (the segment your product and geography can actually serve) and SOM (what you can realistically win in the next few years) — but build the number bottom-up. Count real customers and multiply by a real price. As a purely hypothetical example: 50,000 clinics that could use the product, at $1,200 a year each, is a $60 million annual market.

Avoid the sentence “if we capture just 1% of a $50 billion market”. It signals that the number came from an analyst report rather than from knowing who the customers are, and every investor has read it hundreds of times.

The question it answers: how does money reach you? Who pays, how much, how often, and through what channel. If you have real pricing and real customers paying it, show that. If you know your unit economics — what it costs to acquire a customer, what they are worth over time, your gross margin — include them. If you do not know yet, show the pricing you are testing and say so. One clear model is better than four possible revenue streams; a list of monetisation options tells an investor you have not chosen one.

The question it answers: is it working? One chart, going up and to the right, with a labelled axis and a time scale. Revenue is the strongest evidence, followed by paying customers, then active usage and retention, then signed pilots and letters of intent, then a waitlist. Use whichever is the most honest high-water mark you have.

Pre-launch companies still have traction to show: the number of customer interviews completed, design partners committed, the speed at which the team ships. What matters is the slope, not the absolute figure. Avoid cumulative charts that can only ever rise, and avoid vanity numbers such as downloads or page views with no sign of what happened next.

The question it answers: why will you win, and what happens when the incumbent notices? Never write “we have no competitors”. It means either that you have not looked or that there is no market. The status quo — the spreadsheet, the agency, doing nothing at all — is always a competitor, and frequently the toughest one.

A simple comparison table or a two-axis map works, provided the axes are things customers care about rather than things chosen so that you land alone in the top-right corner. Then state your unfair advantage plainly: a distribution channel, proprietary data, a cost structure, or a piece of insight the incumbents are structurally unable to act on.

The question it answers: how do the next hundred customers find you? Name one or two channels you have actually tested, with the early numbers from each: response rates, conversion, cost per customer. “Social media, content, partnerships, SEO and events” is a list of every channel that exists, not a plan. A narrow, evidence-backed wedge — one customer type, one channel, one repeatable motion — is far more convincing than breadth.

The question it answers: do you understand the economics of your own business? Show three years of projections at a summary level — revenue, costs, burn, headcount — and the handful of assumptions that drive them. Nobody believes year three, and nobody is supposed to. The slide is a test of reasoning: do the assumptions connect to the traction and go-to-market slides that came before it? Keep the full model in a spreadsheet, ready to send the moment someone asks.

The question it answers: what do you want, and what will it buy? State the amount you are raising, the runway it provides (18 to 24 months is the conventional target), and the specific milestones the money gets you to — ideally the ones that make the next round possible. Break the use of funds into three or four buckets. “We're raising $1.5M to reach $80K in monthly recurring revenue within 18 months” is an ask. “We are open to discussing investment” is not.

End with your contact details on the same slide. A closing slide that only says “Thank you” wastes the one slide that stays on screen the longest during questions.

Y Combinator's advice on pitch deck design reduces to three words: make every slide legible, simple and obvious. A deck is not a design portfolio. Its visual job is to get out of the way of the idea.

• One idea per slide. If a slide makes two points, it is two slides.
• Write headlines as conclusions. “Revenue grew 22% a month for six months” instead of “Traction”. A reader who only sees your headlines should still get the whole pitch.
• Use large type. Guy Kawasaki's 10/20/30 rule — ten slides, twenty minutes, nothing smaller than thirty-point font — is deliberately blunt. If the text has to shrink to fit, there is too much text.
• Send a PDF. It opens on every device, looks identical everywhere, and cannot break a font or an animation. Keep the file small enough to email.
• Keep the numbers consistent. If the traction slide says 1,200 users and the financials imply 900, the investor stops trusting both.
• Be plain before being pretty. High contrast, one typeface, one accent colour, and your logo small in a corner rather than large on every slide.

• Too long. Thirty slides means you have not decided what matters. Put the detail in an appendix and send it on request.
• No ask. A surprising number of decks never state how much is being raised.
• Jargon before clarity. If a smart person outside your industry cannot explain what you do after the third slide, rewrite the first three.
• Hiding the weakness. Every company has one. Investors will find it; a founder who names it first and explains the plan reads as credible, not weak.
• Mass-mailing one deck. The 40-meeting average from the DocSend study is a reason to research each investor, not a reason to spam. Check that they invest at your stage, in your sector and in your geography before you send anything.
• Polishing the deck instead of the business. A deck can only report what is true. When the slides are hard to write, the usual cause is that the traction, not the wording, needs the work.

• Can a stranger say what the company does after ten seconds on the solution slide?
• Do the headlines alone, read in order, tell the complete story?
• Does the product video open on a phone, without a login, in under three seconds?
• Does every number appear the same way everywhere it is used?
• Is the amount you are raising stated explicitly, with the milestones it buys?
• Has someone outside the company read it cold and explained it back to you?
• Is it a PDF, sensibly named — “Company – Seed Deck – Sep 2026.pdf”, not “deck_final_v7.pdf”?

Between ten and fifteen for the main deck. Guy Kawasaki argues for exactly ten; Sequoia's template has ten sections; the twelve here add a product video and a go-to-market slide. The decks in the DocSend study averaged 19 pages, but those founders were competing for under four minutes of attention, so shorter is safer. Anything beyond the core story belongs in an appendix.

Both are legitimate. Sequoia's template places the team eighth, after the business is explained. At pre-seed and seed, where there is little traction to evaluate, putting the team first makes sense because the founders are the main thing being assessed. The underlying rule is simple: open with your strongest evidence. If that is revenue growth, lead with traction; if it is an unusually qualified team, lead with the team.

Sixty to ninety seconds. Long enough to show one user completing the core job from start to finish; short enough to be watched in full inside a deck that gets under four minutes in total. Keep a longer walkthrough ready for the follow-up meeting, where the investor has already decided to pay attention.

Usually not. State the amount you are raising and, if relevant, the instrument, but leave valuation for the conversation. A number in a widely forwarded PDF anchors every negotiation before you know how much interest there is, and it goes out of date the moment a term sheet arrives. If a lead investor has already set terms, it is reasonable to include them.

Yes, if asked — refusing mostly just costs you the meeting. Send the send-ahead version, the one that reads clearly without narration, as a PDF or a tracked link. Do not rely on a confidentiality agreement; most investors will not sign one for a deck, so keep anything truly sensitive out of it and share it later in due diligence.

“Lessons From A Study of Perfect Pitch Decks: VCs Spend An Average of 3 Minutes, 44 Seconds On Them,” TechCrunch, June 8, 2015, https://techcrunch.com/2015/06/08/lessons-from-a-study-of-perfect-pitch-decks-vcs-spend-an-average-of-3-minutes-44-seconds-on-them. DocSend and Tom Eisenmann (Harvard Business School), “What We Learned From 200 Startups Who Raised $360M,” 2015, https://www.slideshare.net/DocSend/docsend-fundraising-research-49480890. Sequoia Capital, “Writing a Business Plan,” https://www.sequoiacap.com/article/writing-a-business-plan/. Y Combinator Startup Library, “How to build your seed round pitch deck,” https://www.ycombinator.com/library/2u-how-to-build-your-seed-round-pitch-deck, and “How to design a better pitch deck,” https://www.ycombinator.com/library/4T-how-to-design-a-better-pitch-deck. Guy Kawasaki, “The 10/20/30 Rule of PowerPoint,” https://guykawasaki.com/the_102030_rule/.