The short answer
A pitch deck is ten to fourteen slides that tell one story in a fixed order: what the company does, the problem, the solution, why now, the market, the product, the business model, the traction, the competition, the team, the financials and the ask.
Investors give a seed deck under four minutes, 3 minutes 44 seconds on average in DocSend’s figure updated in March 2026, and 58% of them reach the last slide, so the first four slides carry the decision and every slide after them has about fifteen seconds to make its point. What follows is the order, what each slide must say, the design rules that survive a glance, and the mistakes that end the read.
The reading data, before the slides
- A seed deck is read for 3 minutes 44 seconds on average and a pre-seed deck for 4 minutes 10 seconds (DocSend’s figures, updated in February and March 2026, as published by a benchmark page read on 7 November 2026). The seed completion rate is 58%: four decks in ten are abandoned before the end.
- The first slide gets about 23 seconds and the slides after it about 15 seconds each (Papermark’s 2024 dataset, same page). A slide that needs a minute to understand is skipped, not studied.
- Of the readers who reach slide four, 82% finish the deck (Storydoc’s 2025 sessions). The decision to keep reading is made on the cover, the problem, the solution and the fourth slide.
- In 2024 investors spent 40% more time on seed-stage team slides and 30% more on pre-seed team slides than the year before, while the seed competition slide received 48% less attention and the pre-seed market size slide 19% less (Dropbox’s DocSend release of 18 December 2024, from more than 400 pre-seed and seed startups).
- The successful seed rounds in that release closed in twelve weeks or less, and pre-seed rounds averaged twelve weeks. A deck is read many times over a quarter; it has to work without you in the room.
The twelve slides, in order
The order below follows Sequoia’s template, which has been the common ancestor of most decks since it was published, with the product and traction slides added where investors now look for them. Each slide is one idea; if a slide holds two, split it.
| Slide | What it must say | The mistake that ends the read |
|---|---|---|
| 1. Company purpose | The company in one declarative sentence: who it is for and what it does for them. | A tagline. “Reimagining how teams work” says nothing a reader can check. |
| 2. Problem | The pain of the customer, how it is handled today and what is wrong with that. Name the person, the cost and how often it happens. | A problem nobody pays to solve, or one described in the abstract. |
| 3. Solution | Why your way is different and why it is compelling. One picture of the product is worth the paragraph. | A list of features with no claim about what changes for the customer. |
| 4. Why now | What changed (a technology, a rule, a behaviour) that makes this possible or necessary now and not five years ago. | Skipping it. It is the slide that answers “why has nobody built this”. |
| 5. Market | Who the customer is and how many of them there are, built from the bottom up: customers you can name times what each pays. | A trillion-dollar figure from a report, multiplied by a percentage you chose. |
| 6. Product | What it looks like and how it is used, in one or two screens with large labels. | A wall of screenshots, a video, or a diagram of the architecture. |
| 7. Business model | Who pays, how much, how often, and what one customer costs to win and serve. | A pricing page pasted in. The reader wants the unit, not the menu. |
| 8. Traction | The numbers with their dates: revenue, customers, usage, retention, whichever you have, and the trend. | Vanity counts (sign-ups, followers) standing in for revenue or retention. |
| 9. Competition | Who else serves the customer, including “doing nothing”, and the plan to win against each. | “We have no competitors.” The reader knows the customer is already coping somehow. |
| 10. Team | Why these people, for this problem: what each has built, sold or lived that bears on it. | Logos of former employers with no line about what the person did there. |
| 11. Financials | The next twelve to twenty-four months: revenue, costs, cash, and the three assumptions that drive them. Sequoia’s rule: if you have any, include them. | A hockey stick with no assumptions, or five-year projections to the dollar. |
| 12. The ask | How much, what it buys (the hires, the months, the milestones) and what the company looks like when it is spent. | An amount with no use of funds, or a valuation on the slide. |
What each of the first four slides has to do
The cover is read for about twenty seconds, so it carries the one sentence from slide one and nothing else: the name, the sentence, the stage and the date. The problem slide is where belief is won or lost; investors who do not believe the problem is real, painful and frequent do not read the solution. Write it as a scene, with the person, the cost and the frequency: “A plumbing company with six vans loses two afternoons a month chasing invoices, and the owner does it after hours.” The solution slide then shows the same scene with the product in it. By the fourth slide, why now, the reader has decided whether the next ten minutes of their day go to you; 82% of those who get there finish.
Design rules that survive a glance
- Legible, simple, obvious. Kevin Hale’s three rules from Y Combinator’s guide to deck design, published in 2015 and still the standard: large type, bold, simple fonts, strong contrast, the text at the top of the slide.
- One idea per slide. Hale: simple ideas are ones that are not intertwined with other ideas. If you have seven points, you have seven slides, not three dense ones.
- Obvious at a glance. Make the idea explicit on the slide rather than something the reader has to infer; investors, in his words, are even more easily distracted than most other people.
- No screenshots as decoration, no screencasts, no videos, no animations, no transitions, no accidental humour. A bulleted list beats a diagram unless the diagram is the point.
- Numbers with dates and units. “$41k MRR in October 2026, up from $28k in July” is a fact; “strong growth” is a hope.
- Ten to fourteen slides for the main deck, with an appendix for everything else. Decks of about ten slides had the highest completion rate in Storydoc’s 2026 data; DocSend’s own guidance allows up to nineteen or twenty pages, which is where the appendix lives.
Seed versus pre-seed: what changes
A pre-seed deck has little traction to show, so the team slide, the problem and why now carry it, and the 2024 shift in attention towards team slides is largest there. A seed deck is read for its traction and its business model: what the first customers pay, what they cost to win, and whether they stay. The order of the slides does not change between stages; what changes is which slides are allowed to be thin. A pre-seed financials slide can be twelve months of costs and a hiring plan; a seed financials slide needs revenue with dates.
The mistakes that end the read
- A market slide that multiplies a report’s total by a share you picked. Build it from customers you can name and what each pays.
- A competition slide that says there are none. The customer is coping today with a spreadsheet, an assistant or a competitor; name it and say why you win.
- An ask without a use of funds. The amount is only meaningful next to the months it buys and the milestones it reaches.
- Twenty-five slides in the main deck. Everything past the twelve goes to the appendix, which exists for the second reading.
- Twelve-point text. If it cannot be read from the back of a room, it cannot be read on a phone, where a growing share of decks are opened.
- A deck that only works with the founder talking. Most readings happen without you; the slides have to carry the story alone.
How Castan drafts the deck for you
Castan builds and runs your company with you, from the first website to the monthly investor update. It is an AI, and the deck is one of the things it drafts. On Financials, under Investor, one click writes the twelve slides above from what the workspace already holds: the business definition from the counselor, the odds assessment with its sources, the statements, the runway and the recurring revenue, the customers and what each paid, the open promises, the people, the latest investor update, and the ask you type. Every number on a slide is copied from the books with its date, and where the workspace holds nothing the slide says so in brackets for you to fill, so the deck never claims a market size or a customer it cannot show. You edit each frame, change the whole with a sentence, share it by a link that counts its opens, and print it to PDF from that page. A draft is one Studio run, the same as a page. How the investor update and the deck work in Castan.
Questions people ask
How many slides should a pitch deck have?
Ten to fourteen in the main deck, one idea each, with an appendix for the rest. Decks of about ten slides had the highest completion rate in Storydoc’s 2026 data, and the seed completion rate overall is 58%, so brevity is what gets a deck read to the end.
What order should pitch deck slides be in?
Company purpose, problem, solution, why now, market, product, business model, traction, competition, team, financials, the ask. The first four decide whether the rest is read: 82% of readers who reach slide four finish.
How long do investors spend on a pitch deck?
Under four minutes: 3 minutes 44 seconds on a seed deck and 4 minutes 10 seconds on a pre-seed deck in DocSend’s figures updated in early 2026, about 23 seconds on the first slide and 15 seconds on each slide after it.
What is the difference between a pitch deck and a business plan?
A pitch deck is the story in twelve slides, read in minutes, written to earn a meeting. A business plan is the operating document behind it: the definition, the customers, the numbers and the assumptions, which is what the deck’s slides are copied from.
Sources
- 1Sequoia Capital: writing a business plan (the pitch deck template)company purpose, problem, solution, why now, market, competition, business model, team, financials, vision
- 2Y Combinator, Kevin Hale: how to design a better pitch deck, 15 November 2015legible, simple, obvious; one idea per slide; no screenshots, videos or transitions
- 3Dropbox DocSend press release, 18 December 202440% more time on seed team slides, 48% less on seed competition slides; rounds closing in twelve weeks; more than 400 startups
- 4Hummingdeck: pitch deck benchmarks 2026, citing DocSend, Papermark and Storydocseed 3:44 and pre-seed 4:10 per deck, 58% completion; 23 seconds on the first slide and 15 after; 82% who reach slide four finish
- 5Storydoc: the state of pitch deck presentations in 2026about ten slides had the highest completion rate; 31% of readers leave in ten seconds
- 6Castan: the pitch deck on Financials, drafted from the workspace
Prices and terms at other companies change over time. Every figure on this page carries the day it was read, and the page it was read on is linked above. If one has moved since, write to hello@castan.ai and the page is corrected. General information, not tax, legal or accounting advice.
