What Investors Read in Your Deck, and How Long You Have

If you have read anything about pitch decks in the last decade, you have read that investors spend an average of three minutes and forty-four seconds on one. It is the most quoted statistic in early-stage fundraising, it comes from DocSend, and it is from 2015.

DocSend's own page on seed decks, last updated in March 2026, still prints it in the present tense with no year, no sample size and no citation attached.¹ That is almost certainly why it keeps circulating as current. It is not current, and the way it has moved since is more useful to a founder than the original figure ever was.

Attention Has Roughly Halved

DocSend has published the same measure repeatedly, and the series is consistent in direction. The 2015 study, covering more than 200 decks from companies that raised 360 million dollars between them, gave 3 minutes 44 seconds.² The 2020 full-year average was 3 minutes 10 seconds. By the third quarter of 2021 it was 2 minutes 44 seconds, with the year to date at 2 minutes 47.³ In the fourth quarter of 2023 it hit what DocSend called an all-time low of 2 minutes 24 seconds.⁴ The most recent figure in a dated press release is 2 minutes 30 seconds, for the first quarter of 2024.⁵ DocSend's own weekly tracker ran lower still, at 2 minutes 18 seconds in the week of 10 June 2024, before the time-spent series stopped.¹²

So the window has shrunk by about a third over a decade, and most of the compression happened after 2020. A deck built on advice calibrated to nearly four minutes of attention is being read in two and a half.

That matters for length, and the length advice has reversed accordingly. In DocSend's own comparison of successful against failed decks, published by its co-founder in 2019, successful decks averaged 20 pages against 17 for decks that failed to raise.⁶ In DocSend's 2023 pre-seed report, covering over 200 pre-seed startups across 2022 and the first half of 2023, successful decks averaged 16 slides and unsuccessful ones 19.⁷ The same organisation, measuring deck length against outcome, found the opposite relationship eight years apart, though the earlier sample covered seed and Series A rounds and the later one pre-seed only, so the two are not strictly comparable. The likely reconciliation is the attention figure above: at nearly four minutes a longer deck could work, and at two and a half it cannot.

The Finding That Inverts the Obvious

The most useful thing in DocSend's research is not the total. It is a comparison of successful decks against failed ones that almost nobody quotes, published by DocSend's own co-founder in 2019.

In decks that succeeded, investors spent 22 seconds on average on the company purpose slide. In decks that failed, they spent an average of 61 seconds on the opening product page. Overall, failed decks absorbed 56 percent of total viewing time on product, financials and team, while successful decks took only 26 percent of viewing time on those same three.⁶

Read that the obvious way and it says nothing. Investors spending longer on your product must be more interested in your product. The data says the opposite. Long dwell time on those sections is a symptom of a reader working to understand something that has not been made clear. Successful decks distribute attention across the whole story; failing decks trap it in one place while the reader tries to work out what the company does.

This is the finding a founder can do most with in the whole body of DocSend research, and it reframes what a founder should want from a deck analytics tool. A slide that holds someone for a minute is not necessarily winning. It may be the slide that lost them.

What Gets Read

DocSend publishes a per-section breakdown for seed decks across twelve sections. Business model draws the most time at 64 seconds, then product at 59, traction at 40, team at 38, financials at 37, competition and problem and solution at 34 each, fundraising ask at 32, market size at 29, company purpose at 26 and "why now" at 23.¹

One caveat on those numbers, and it is the kind that gets missed. Those twelve figures sum to seven and a half minutes, which is roughly double the 2015 headline printed on the same page and triple the current figure. They cannot be a decomposition of a single reader's session. They are more likely averages conditional on a section being reached at all, or drawn from a different sample. So use them as a ranking of relative weight, which is what they reliably show, and not as a budget of where three minutes go.

Reading them as a ranking still says something. Business model and product take almost twice the attention of company purpose, and roughly half of what a founder agonises over at the front of the deck is skimmed.

DocSend also reports that only 58 percent of decks are viewed to completion, so four in ten decks are abandoned before the last slides.¹ That page gives no year or sample for the figure, which is worth knowing before repeating it. What is dated and specific is the 2020 pre-seed comparison: successful decks held investors an average of 4 minutes 10 seconds against 1 minute 36 for decks that did not raise, with some sessions lasting only 20 to 30 seconds, which is one or two slides in.⁸

There is a related thing the research does not say, and it is worth resisting. There is no published DocSend finding on whether investors read decks in order or skip around. The analytics exist internally; the finding has not been published. Anyone telling you investors jump straight to the team slide is asserting something the data has not shown.

The One Finding Two Independent Methods Agree On

Where deck research gets most reliable is where it stops being deck research. In a randomized field experiment on AngelList, Shai Bernstein of Harvard Business School, Arthur Korteweg of USC Marshall and Kevin Laws of AngelList randomised which information about fundraising startups investors saw. They emailed 4,494 active investors across 16,981 emails and 58 versions, featuring 21 early-stage startups over an eight-week period. Revealing information about the founding team raised click rates by 2.6 percentage points against a 16.5 percent baseline, roughly a 16 percent relative increase. Information about traction and about existing lead investors produced no significant response.⁹

The nuance in that result is important and usually dropped. The null on traction is an average across all investors. Less experienced investors in the experiment responded to every category of information; the team-only pattern is driven by the experienced ones. So the honest statement is that experienced early-stage investors moved on team information and did not move on traction, not that traction is irrelevant.

An entirely different method reaches the same place. Gompers, Gornall, Kaplan and Strebulaev surveyed 885 institutional venture capitalists at 681 firms. The management team was named as an important factor by 95 percent of firms and as the single most important factor by 47 percent, against 37 percent naming business-related factors as most important.¹⁰ A randomized experiment on nearly four and a half thousand angels and a survey of 885 institutional investors, using nothing in common methodologically, land on the same conclusion.

DocSend's own 2024 data moved in the same direction. Analysing more than 400 pre-seed and seed startups, investors spent 40 percent more time on seed-stage team slides and 30 percent more on pre-seed team slides than the year before, while pre-seed market size drew 19 percent less attention and seed competition 48 percent less.¹¹

The Uncomfortable Context

Here is the part that should temper all of the above. In that same survey of 885 venture capitalists, only 10 percent of deal flow arrived inbound from company management. Over 30 percent came through professional networks, 20 percent from referrals by other investors, 8 percent from portfolio companies and almost 30 percent was proactively self-generated.¹⁰

Roughly nine deals in ten never start with someone sending a deck. The deck is not the acquisition channel; it is what survives or fails a meeting you already have. Polishing it is worth doing and it is not a substitute for being in the flow of referrals, which is a slower and less satisfying problem to work on.

Two other caveats belong on the DocSend material specifically, and we would rather state them than let the numbers carry more weight than they can hold. The sample is founders who chose a paid deck-tracking product, which skews toward the more process-sophisticated and toward US technology companies. Time on page measures a browser tab being open rather than attention, so an investor who opens a deck and takes a phone call inflates the average. Successful and unsuccessful are labelled after the fact, so none of it is causal.

And the research appears to have stopped. DocSend's live pitch deck metrics tracker has no data point after September 2024, its last research press release is dated December 2024, and its blog through mid-2026 covers data rooms and M&A rather than fundraising.¹² The most recent deck-behaviour data available is about two years old, and the most quoted number in the field is eleven.

What To Do With This

Build for two and a half minutes, not four. Sixteen slides is a better starting point than twenty at pre-seed.

Put the company purpose first and make it survivable in twenty seconds. The successful decks in the 2015 data were the ones where the reader understood the business quickly and then moved on, not the ones that held attention longest at the front.

Treat long dwell time on product or financials as a warning rather than a win, and go and fix whatever in that section is making someone read it twice.

Make the team section carry weight, because it is the one thing a randomized experiment and a survey of 885 investors independently agree on, and because DocSend's most recent data shows readers spending more time there than they used to.

Then spend the rest of your effort on the nine deals in ten that never start with a deck. On AngelHive that means the profile a network member sees before any deck is opened: who the founders are, what has been checked, and who else has looked. That is what produces the meeting, and the deck only has to survive it.

Sources

1. DocSend, What VCs really want to see inside your seed deck, Justin Izzo, originally published 3 May 2022, updated 10 March 2026. [https://www.docsend.com/blog/what-vcs-really-want-to-see-inside-your-seed-deck/](https://www.docsend.com/blog/what-vcs-really-want-to-see-inside-your-seed-deck/)

2. TechCrunch, Lessons From A Study of Perfect Pitch Decks: VCs Spend An Average of 3 Minutes, 44 Seconds On Them, 8 June 2015, reporting DocSend's 2015 fundraising study. [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/](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/)

3. DocSend via PR Newswire, DocSend Pitch Deck Interest Q3 2021 Data Shows Consistent Highs for VC Interest, Record Lows for VC Time Spent on Pitch Decks, 19 October 2021. [https://www.prnewswire.com/news-releases/docsend-pitch-deck-interest-q3-2021-data-shows-consistent-highs-for-vc-interest-record-lows-for-vc-time-spent-on-pitch-decks-301402816.html](https://www.prnewswire.com/news-releases/docsend-pitch-deck-interest-q3-2021-data-shows-consistent-highs-for-vc-interest-record-lows-for-vc-time-spent-on-pitch-decks-301402816.html)

4. DocSend via PR Newswire, DocSend 2023 Year-End Data Indicates Positive Fundraising Momentum Going Into 2024, 17 January 2024. [https://www.prnewswire.com/news-releases/docsend-2023-year-end-data-indicates-positive-fundraising-momentum-going-into-2024-302036345.html](https://www.prnewswire.com/news-releases/docsend-2023-year-end-data-indicates-positive-fundraising-momentum-going-into-2024-302036345.html)

5. Dropbox DocSend via PR Newswire, Investor Activity Surpasses 2021 Engagement Levels, Hits Q1 Record High According to DocSend 2024 Data, 11 April 2024. [https://www.prnewswire.com/news-releases/investor-activity-surpasses-2021-engagement-levels-hits-q1-record-high-according-to-docsend-2024-data-302113696.html](https://www.prnewswire.com/news-releases/investor-activity-surpasses-2021-engagement-levels-hits-q1-record-high-according-to-docsend-2024-data-302113696.html)

6. TechCrunch, Data tells us that investors love a good story, Russ Heddleston, co-founder and chief executive of DocSend, 12 April 2019, describing the 2015 analysis. [https://techcrunch.com/2019/04/12/data-tells-us-that-investors-love-a-good-story/](https://techcrunch.com/2019/04/12/data-tells-us-that-investors-love-a-good-story/)

7. DocSend via PR Newswire, From Growth-At-All-Costs to Cost of Missing Out: DocSend Pre-Seed Report Shows Investor Shift to Long-Term Profitability and Risk Aversion, 24 August 2023. [https://www.prnewswire.com/news-releases/from-growth-at-all-costs-to-cost-of-missing-out-docsend-pre-seed-report-shows-investor-shift-to-long-term-profitability-and-risk-aversion-301908433.html](https://www.prnewswire.com/news-releases/from-growth-at-all-costs-to-cost-of-missing-out-docsend-pre-seed-report-shows-investor-shift-to-long-term-profitability-and-risk-aversion-301908433.html)

8. DocSend, How To Structure A Pre-Seed Pitch Deck: Beyond The "Minimum Viable PowerPoint", reporting 2020 pre-seed data. [https://www.docsend.com/blog/how-to-structure-a-pre-seed-pitch-deck-beyond-the-minimum-viable-powerpoint/](https://www.docsend.com/blog/how-to-structure-a-pre-seed-pitch-deck-beyond-the-minimum-viable-powerpoint/)

9. Shai Bernstein, Arthur Korteweg and Kevin Laws, Attracting Early-Stage Investors: Evidence from a Randomized Field Experiment, The Journal of Finance, volume 72, issue 2, April 2017, pages 509-538. [https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2432044](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2432044)

10. Paul Gompers, Will Gornall, Steven N. Kaplan and Ilya A. Strebulaev, How Do Venture Capitalists Make Decisions?, NBER Working Paper 22587, later published in the Journal of Financial Economics, volume 135, issue 1, 2020, pages 169-190. [https://www.nber.org/system/files/working_papers/w22587/w22587.pdf](https://www.nber.org/system/files/working_papers/w22587/w22587.pdf)

11. Dropbox DocSend via PR Newswire, VCs prioritize people in an AI-heavy landscape, according to new Dropbox DocSend report, 18 December 2024. [https://www.prnewswire.com/news-releases/vcs-prioritize-people-in-an-ai-heavy-landscape-according-to-new-dropbox-docsend-report-302334444.html](https://www.prnewswire.com/news-releases/vcs-prioritize-people-in-an-ai-heavy-landscape-according-to-new-dropbox-docsend-report-302334444.html)

12. DocSend, Pitch Deck Interest Metrics, live tracker, most recent data point 23 September 2024. [https://www.docsend.com/pitch-deck-metrics/