How to design a pitch deck that reads in 90 seconds

I design a pitch deck for a ninety-second read by putting the decision first, giving each slide one claim, and making the headlines carry the argument. I keep only the evidence needed to judge that claim, then test the exported deck without narration.

Pitch deck slide order for a ninety-second read

I treat ninety seconds as a design constraint, not a promise about attention spans. With ten slides, that allows nine seconds per slide on average, including transitions. The reader needs to understand the offer, believe there is a reason to care, and know what decision comes next. Anything else has to justify its space.

Sequoia's business-plan guidance is a useful content check: explain the business, customer problem, timing, market, commercial logic and people behind it. I use those subjects to check completeness, not as a compulsory slide order. For an early-stage funding deck, I would start with this sequence:

• Offer: name the customer and the result the product delivers.
• Problem: show the specific cost of the current approach.
• Timing: explain what has changed to make adoption plausible.
• Product: demonstrate the mechanism that changes the customer's situation.
• Evidence: show usage, revenue or a clearly labelled validation result.
• Market: define the reachable customer group and its spending.
• Alternatives: explain why customers would switch from their current choice.
• Economics: show how the business earns money and serves customers.
• Team: connect relevant experience to the work still ahead.
• Ask: state the amount sought and the milestone it funds.

How to keep one idea per pitch deck slide

I write the slide's claim before choosing a layout. 'Product' is a filing label. 'One approval replaces three separate handovers' is a claim the reader can assess. The supporting material should explain or substantiate that sentence. If half the slide proves something else, I split it or remove the weaker point.

One idea doesn't mean one object. A claim may need a chart, a short annotation and a source line. Those elements belong together when they answer the same question. A product screenshot beside a hiring plan usually doesn't. Shrinking both until they fit preserves the content but destroys the reading order, leaving the reader to decide which point matters.

I check the argument by reading only the headlines in sequence. They should describe a coherent business case without the body copy. If two consecutive headlines say effectively the same thing, one slide probably hasn't earned its place. I keep wording specific enough to disagree with, but I don't turn an ambition into an achieved result. 'Designed to reduce processing time' and 'Reduced processing time by 30%' require different evidence.

How much text should a pitch deck slide have?

For a deck intended to circulate without a presenter, I start with a headline of roughly six to twelve words and twenty to forty words of supporting copy. These are working limits, not research findings. A financial slide may need more labels. A product demonstration may need fewer. The useful test is whether every sentence changes the reader's understanding.

I write for an absent narrator. A chart labelled 'Growth' leaves too much work to the reader. A headline naming what grew, plus a visible period and unit, makes the evidence interpretable. When a claim depends on a qualification, I place that qualification beside it. I don't hide 'pilot customers only' in a footnote that becomes illegible after export.

I move definitions, detailed assumptions, secondary use cases and full research notes into an appendix. That isn't permission to conceal material weaknesses. If customer concentration changes the meaning of a revenue claim, it belongs with the revenue. My filter is simple: remove information that delays the decision, retain information that could reasonably change it. The main deck should remain honest even if nobody opens the appendix.

Pitch deck design tips for layout and typography

I build three layouts before styling individual slides: a claim with evidence, a comparison, and a single visual with an annotation. Repeating them reduces the work of finding the headline, proof and caveat each time. I keep margins consistent and align related objects. A slide shouldn't need a new visual grammar simply because its subject has changed.

For a widescreen deck, I might start with 36-point headlines, 24-point body text and 16-point source notes, then inspect the exported file at ordinary laptop size. Those sizes are starting values, not guarantees of readability. Typeface proportions, contrast and line length matter. If the smallest note carries essential meaning, I promote it into the main content rather than asking the reader to zoom.

Pentagram's Mastercard identity work offers a useful reference for visual economy: the overlapping circles and a limited colour vocabulary carry a recognisable identity. I borrow that principle of restraint, not its appearance. In a deck, I use one type family, a neutral background and one accent colour with a defined purpose. I skip decorative icons and background textures unless they communicate something the content cannot.

How to make pitch deck charts readable

I choose the chart around the comparison the reader needs to make. For change over time, I usually start with a line. For differences between categories, I start with bars. I avoid a dashboard compressed into a slide: six small charts create six competing tasks. One legible comparison is more useful than an inventory of everything the company measures.

Consider a hypothetical business whose monthly revenue rose from £20,000 to £32,000 over six months. I would label the series directly, show the period and currency, and state whether the figures are recognised revenue or another measure. I wouldn't imply that revenue growth proves retention. If the slide's claim is about repeat purchasing, the evidence needs to show repeat purchasing instead.

I distinguish actual results from forecasts using both labels and styling, so colour isn't the only cue. Bar charts normally start at zero; any departure needs a clear reason and visible treatment. I remove ornamental gridlines but retain the scale, source and necessary context. For a market estimate, I show the customer count and spending assumption rather than presenting a large number without its calculation.

How to test a pitch deck in ninety seconds

I export the deck before testing it. Font substitution, compressed screenshots and awkward line breaks can undermine a careful layout. I open the file on a laptop and a phone, check reading order, and confirm that links work. I also check that the filename, date and contact details identify the right version. These are small production decisions with avoidable consequences.

I give someone unfamiliar with the business ninety seconds to read it without commentary. Afterwards, I ask them to describe the customer, the problem, the product's difference, the strongest evidence and the ask. I record what they remember accurately, what they misunderstand and where they stop. This is a practical diagnostic, not a statistically meaningful study. One reader can reveal a problem without representing the whole audience.

I revise in order: sequence, claims, evidence, then styling. If the reader remembers the illustration but cannot explain the offer, I haven't solved the communication problem. If they understand the offer but miss the ask, I make that decision more explicit. I repeat the test with a fresh reader rather than coaching the first one until they know what I meant.

Questions people ask

How many slides should a pitch deck have?

I start with ten slides for a short funding deck, then adjust to the decision and available evidence. Ten isn't a rule: I would rather use twelve readable slides than eight crowded ones.

What is the best font for a pitch deck?

I choose a typeface with clear letterforms, useful weights and reliable export support. Arial or Inter can work; I check the actual deck at reading size rather than choosing by reputation.

Should I send a pitch deck as a presentation or a document?

I usually send a fixed-layout document so recipients see the intended typography and spacing. I keep an editable presentation separately and check the exported file before sending it.

Can I use the same pitch deck for presenting and emailing?

I keep the same argument but often make two versions. The emailed deck needs enough explanation to stand alone; the presented version can leave more detail to speech.

Where I checked my thinking

• Sequoia Capital: Writing a business plan
• Pentagram: Mastercard identity