Section 1
Key takeaways
• Founders systematically over-build. DocSend's research finds the typical pitch deck runs roughly 20 pages with about 50 words a slide, while investors now give it under three minutes, production volume vastly outruns available attention. • Shorter completes better. Storydoc's behavioral data shows sales decks average a 22% completion rate, rising to 32% when kept under 10 slides, and the pattern holds at the bottom of the funnel, where proposals climb from 38% to 43%. • The job is to earn time, not fill space. DocSend's outcome data ties results to engagement, not page count: decks that raised money held attention four-plus minutes; decks that failed got about ninety seconds. • Concision is a purchase driver, not a stylistic preference. Gartner finds that giving buyers concise, decision-enabling information makes them 2.8 times more likely to close a high-quality, low-regret deal. • The fix is a forcing function, the Subtraction Pass, that makes substance earn its place on one page before any design budget is spent.
Section 2
The deck you built is not the deck they read
Start with the gap between what founders produce and what buyers consume, because every other mistake compounds from there. On the supply side, the over-build is well documented. DocSend's pitch-deck research, drawn from how investors actually open and scroll documents, found that most pitch decks run roughly 20 pages, with about 50 words on a slide. That is a serious artifact: twenty pages, a thousand words, hours of design, version after version. It feels like diligence. It reads like commitment. On the demand side, the attention has collapsed. The same DocSend research notes investors now spend under three minutes on a deck. PitchGrade, citing DocSend's data, puts the precise first-read benchmark at three minutes and forty-four seconds, the entire window in which a quantified outcome has to land. (A quantified outcome is the result you produce stated as a number a buyer cares about, revenue, time, risk, not a description of what you do.) Three minutes and forty-four seconds against twenty pages is not a reading; it is a triage. The reader is scanning for a reason to keep going, and most of your slides never get that reason. Here is the part founders miss: the volume is not neutral. It is not that the extra slides are simply ignored. They actively dilute the few that matter, because attention is a fixed budget being spread across more surface. A practitioner framing captures it well, a 25-slide deck dilutes the four slides that decide the outcome; a 12-slide deck forces you to choose them. Every page you add is a page of attention you subtract from the page that closes. And the stakes are visible in the outcomes, not just the engagement. DocSend's data ties results to attention held, not to page count: investors spent more than four minutes on the decks that successfully raised money, while the decks that failed got about ninety seconds. Read that carefully, the difference between a raise and a pass was roughly two and a half minutes of held attention. The job was never to present more. It was to earn time. The deck that earns four minutes wins; the deck that loses the reader at ninety seconds loses regardless of how many slides sat unread behind the exit. So the first reframe is this: you are not building a document, you are competing for a finite block of attention, and most of what you build never enters the contest.
Section 3
Shorter completes better, and the data is not opinion
The instinct, once a founder accepts the attention gap, is to assume "shorter" is a matter of taste, that some buyers like long, some like short, and it all averages out. The behavioral data says otherwise. Storydoc's analysis of 1.3 million reading sessions across more than 500,000 unique presentations, published in November 2025, is the largest first-party look at this behavior available, not a survey of preferences, but a record of what readers actually did. The headline is unambiguous. Completion rate, the share of readers who reach the last slide, averages around 22% for sales pitch decks, and rises to 32% when the deck is kept under 10 slides. Trimming slides lifted completion by roughly 45% in relative terms. Same content category, same buyers; the shorter format simply got finished more often. The objection writes itself: of course top-of-funnel skimmers bail on long decks, but surely serious, late-stage buyers read everything. They do not. The same Storydoc dataset shows the pattern holding at the bottom of the funnel, where intent is highest. Proposals, documents a buyer has explicitly asked for, average 38% completion, the strongest of any category. Even there, decks with fewer than 10 slides hit 43%. Shorter won where the buyer most wanted to read. That is the tell that this is structural, not situational: brevity is not a concession to distracted prospects, it is a better fit for how committed ones read too. "Ten slides. This is the optimal number of slides in a PowerPoint presentation because a normal human being cannot comprehend more than ten concepts in a meeting.", Guy Kawasaki, The Art of the Start, on his 10/20/30 rule Kawasaki's rule predates the data by years, but the data validates the instinct behind it: there is a ceiling on how many concepts land in one sitting, and decks that respect the ceiling get finished. One honest caveat keeps this from tipping into dogma. "Shorter" is not "amputated." Some practitioners point out that decks in the 11-to-20-slide range can do well when the extra slides carry genuine substance a buyer needs to decide, and that a review often runs a few minutes, not seconds, for a deck that earns it. The lesson is not a slide quota. It is that every slide must earn its place against the attention it costs. Ten strong slides beat twenty padded ones; ten amputated slides beat nothing. The discipline is subtraction with a floor, not minimalism for its own sake, which is exactly the distinction the one-page business case is built to enforce.
Section 4
The BGA framework: The Subtraction Pass
Here is the operating model. The Subtraction Pass is a forcing function you run before design, not a cleanup you run after. The premise: design is the most expensive way to discover your argument is thin. Find out on one page, for free, first. Run it in four steps, each with a metric that tells you whether you passed. 1. State the outcome as one number. Before anything else, write the single quantified result a buyer gets, in one sentence, in money or time or risk. Not "we streamline operations", "we cut your onboarding from 14 days to 3." Metric: can a stranger read it once and repeat it back accurately? If they paraphrase it into mush, the number is not sharp enough yet, and no slide will sharpen it. This is the move from describing your work to pricing your buyer's result, the same translation discipline behind a features-vs-outcomes value map. 2. Reduce to the four load-bearing claims. Every persuasive deck rests on exactly four claims: the problem is real and expensive, your outcome solves it, the proof is credible, and the next step is low-risk. Write each as one line. Metric: remove any one of the four, does the argument collapse? If it survives the removal, that line was decoration, not structure. Cut it. Most twenty-page decks are four load-bearing claims wearing sixteen pages of costume. 3. Fit the whole argument on one page. Put the outcome and the four claims on a single page, no design, no template, plain text. This is the floor the one-page business case sets, and it is deliberately unforgiving. Metric: does it hold together read top to bottom in under sixty seconds, by someone who has never heard your pitch? If you need to narrate it for it to make sense, the page is not done. The page has to win without you in the room, because increasingly it is read without you in the room. 4. Design only what survives. Now, and only now, you build slides, one per surviving claim, plus the outcome and the ask. Design serves the substance that already proved it could stand alone. Metric: time-on-deck and completion, the numbers Storydoc and DocSend track. You are not optimizing for how the deck looks in a portfolio; you are optimizing for whether a stranger finishes it and remembers the number. The reason this works is not aesthetic, it is commercial. Gartner's buyer research, carried via Prospeo, found that giving buyers concise, decision-enabling information makes them 2.8 times more likely to close a high-quality, low-regret purchase, the kind they do not unwind later. (Prospeo attributes the figure to Gartner without a sample size on the page, so treat the multiplier as directionally strong rather than precise.) Concision is not how you look serious. It is how you make a buyer's decision feel safe enough to make. The Subtraction Pass is just the mechanism that produces it on purpose instead of by accident.
Section 5
What this looks like on a real deck
Take a service founder selling a $40,000 operations engagement, with a 22-slide deck: a company-history slide, a values slide, a six-logo "our stack" slide, three slides of methodology, a team grid, and somewhere in the middle, the one slide that says clients cut fulfillment costs by 30%. Run the Subtraction Pass. The outcome, "we cut your fulfillment cost by 30%, roughly $18,000 a month at your volume", was on slide 11, in eight-point type, under a stock photo. It moves to the front and becomes the spine. The four load-bearing claims survive: the cost is real, the 30% is the fix, two named case results are the proof, a paid two-week pilot is the low-risk next step. Everything else, history, values, the logo wall, two of the three methodology slides, fails the removal test. Pull any of them and the argument does not move. They were costume. The redesigned deck is nine slides. Completion is supposed to rise; per Storydoc's pattern, that is what fewer slides buy. But the deeper change is what the founder learned on the one-page draft: that the methodology slides they were proudest of did nothing for the buyer, and the number they buried was the entire pitch. Design would never have surfaced that. Subtraction did, in an afternoon, before a dollar of design was spent. The same logic governs the live conversation, not just the document. The instinct to walk a prospect through every feature is the demo equivalent of a twenty-page deck, and it loses for the same reason, which is why death by feature tour fails where a tight, outcome-led walkthrough wins. If you have to send something asynchronous, the 60-second Loom is the Subtraction Pass in video form: one outcome, one reason to reply, nothing padded. And when the buyer's real question is "what does this return," you are no longer defending slides, you are defending an ROI you can defend, which is a far stronger place to stand.
Section 6
Reframe honesty: where one-page discipline fails
This is not a universal law, and selling it as one would be the kind of hype this framework exists to strip out. Name the limits. First, complexity floors exist. A deep-tech raise, a regulated-industry proposal, or an enterprise security review genuinely needs pages a starter deck does not, the buying committee will ask for them. The Subtraction Pass does not delete those pages; it relegates them. They live in an appendix the reader reaches only if they want to, not in the main path that competes for the first three minutes. One page up front, depth on demand behind it. Second, "short" can be used as an excuse to be vague. A founder who cannot articulate their outcome will produce a short deck that is short because it is empty, not because it is edited. The metric in step one, a stranger can repeat the number, is the guard against this. Brevity that survives that test is editing. Brevity that fails it is just a thinner version of having nothing to say. Third, the channel matters. A deck read live, with you narrating, can carry more than a deck sent cold to be read alone. But the trend runs against narration: more decks are forwarded, opened after hours, and scrolled without you there. Building for the unattended read is building for where the attention is going, not where it was. Knowing which of these applies to you is the difference between disciplined subtraction and amputation. The framework is a scalpel, and a scalpel in the wrong hands still does damage.
Section 7
What this means
For pre-revenue and early-stage founders. You have the most to gain and the least sunk cost. Build the one-page draft before you ever open a slide tool. You will likely discover your outcome is not yet a number, and that discovery, made early, is worth more than any template. Do not design your way past a weak argument; you will only make the weakness expensive. For five- and six-figure service founders. You already have a deck, and it is almost certainly over-built, because it grew by accretion, a slide added per objection, per quarter, per "we should mention." Run the removal test on every slide this week. The four that survive are your real pitch; the rest are tax on your buyer's attention. Cutting them is not losing information, it is recovering the minutes that close deals. For seven-figure founders and operators. The risk at your stage is institutional bloat, brand guidelines, legal lines, the slide every department insisted on. Make the Subtraction Pass a gate in your sales process, not a one-time cleanup. No deck or proposal goes out until its one-page core has been signed off by someone who was not in the room when it was built. That single checkpoint protects completion rates across every rep, not just the founder's own deck.
Section 8
You're running the Subtraction Pass right when…
You are running it right when a stranger can read your one-page core in under a minute and repeat the outcome number back to you correctly, and when removing any surviving slide visibly breaks the argument. You are running it right when design starts after the substance is proven, not as the method for discovering whether substance exists. You are not running it when your deck has grown a slide every quarter and lost one never; when your strongest number is buried past slide ten; when you are polishing layout to compensate for an argument you have not yet made survive on plain text. If the deck only makes sense with you narrating it, you have not built an argument, you have built a dependency. Strip it back to the page that wins without you, and design up from there. Want the one-page templates and the removal-test worksheet in a single kit? The ConvertOS template pack packages the Subtraction Pass as a fill-in workflow, and the ConvertOS playbook installs it across your whole sales motion. --- If your decks are growing faster than they convert, the fix is a system, not another redesign. The ConvertOS playbook installs the Subtraction Pass across your sales motion. See how the layers fit together at /system, or take it to your own pipeline on a strategy call at /book.
Section 9
Sources
1. Storydoc, "Sales and Marketing Presentations Statistics (Updated 2025)", https://www.storydoc.com/blog/presentation-statistics. Primary first-party behavioral dataset: 1.3 million reading sessions across 500,000+ unique presentations, published November 2025. Source of the completion-by-slide-count figures (sales decks 22%→32%; proposals 38%→43% under 10 slides). Strongest data hook because it measures behavior, not stated preference. 2. Dropbox / DocSend, "How DocSend data helps founders build better pitch decks", https://www.dropbox.com/resources/docsend-pitch-deck-research. DocSend's pitch-deck research. Source of the over-build benchmark (~20 pages, ~50 words/slide; under three minutes of attention) and the outcome split (4+ minutes for decks that raised, ~90 seconds for decks that failed). 3. PitchGrade, "What Investors Actually Read in Your Pitch Deck (DocSend Data)", https://pitchgrade.com/blog/what-investors-read-pitch-deck-docsend-data. Secondary citation of DocSend's headline figure: average investor spends 3 minutes 44 seconds on a deck the first time. 4. Prospeo, "7 B2B Sales Best Practices Backed by 2026 Data", https://prospeo.io/s/b2b-sales-best-practices. Carries the Gartner "concise, decision-enabling information increases ease of purchase by 2.8x" figure. Attributed to Gartner but with no sample size on the page; treated as directional. 5. Think Insights (citing Guy Kawasaki, The Art of the Start), "The 10/20/30 Rule of PowerPoint", https://thinkinsights.net/consulting/10-20-30-rule-presentation. Source of the named-framework quote. Figure-source note: The 2.8x ease-of-purchase multiplier is attributed to Gartner via Prospeo, which does not publish a sample size for that figure; treat it as directionally strong rather than precise. All other figures are drawn from named first-party datasets (Storydoc, DocSend) and were verbatim-confirmed against the cited URLs. Time-sensitive data points carry their publication dates inline. Spot-check load-bearing figures before heavy reuse, as benchmarks shift year to year.