• Twenty-three pitch decks that closed capital.
  • I. Airbnb. The discipline of a clean problem.
  • II. Uber. Selling a future that does not yet exist.
  • III. Tinder. When the product is the pitch.
  • IV. Brex. What to do when traction can carry the deck.
  • V. Notion. Taste as a commercial asset.
  • Eleven habits common to the entire survey
  • Five errors that disqualify an otherwise strong deck
  • Diagnose your own deck

§Twenty-three pitch decks that closed capital.

Every founder asks the same question before their first raise — what does a pitch deck that actually works look like? The honest answer is that it looks like a great many different things, and the founders who study only one example tend to produce decks that resemble a costume more than a document.

What follows is a survey of twenty-three decks that closed institutional capital. Each is annotated with the single thing it does that you should take seriously.

§I. Airbnb. The discipline of a clean problem.

Seed Round, MMVIII · Sequoia · $600,000 · 10 slides

The most circulated pitch deck in living memory, and properly so. Slide two consists of a single sentence: price is an important concern for customers booking travel online. Slide three: hotels leave you disconnected from the city and its culture. The problem is so plainly stated that the solution lands harder when it arrives.

Take seriously: Spend more rhetorical energy on the problem slides than the solution. A stranger reading only your first three slides should already be nodding.

§II. Uber. Selling a future that does not yet exist.

Pre-launch · 25 slides

The original deck was titled UberCab. It reads less like a modern pitch deck and more like a business plan. The deck’s most durable strength is its vision section, which renders a city without the inefficiency of street-hailed cabs as something inevitable.

Take seriously: A vision slide that makes the future tangible outperforms abstract market data.

§III. Tinder. When the product is the pitch.

Seed · 10 slides · Visual format

Tinder’s original deck was unusually visual — screenshots of the product occupy most of every slide. The interface itself does the value-proposition work.

Take seriously: If your product is visual or experiential, your deck should be too.

§IV. Brex. What to do when traction can carry the deck.

Series B · $125M · 19 slides

A master class in numbers. Every slide contains a chart. Growth curves are unambiguous.

Take seriously: When traction can do the work, let it. Reorder your deck so numbers appear earlier than the canonical sequence suggests.

§V. Notion. Taste as a commercial asset.

Early raise · 13 slides

The deck resembled a Notion page. The artifact itself functions as a demonstration of the product’s design philosophy.

Take seriously: Your deck’s design is itself a product demonstration.

§Eleven habits common to the entire survey

1. Length is short. Eighteen of twenty-three decks run 10–15 slides.
2. Problem precedes solution. Every single deck.
3. One number anchors the document. Find yours before you write.
4. Team slide is a closer, not an opener — slide 8 or 9.
5. The ask is specific: amount, allocation in percentage buckets, milestone.
6. Vision is rendered, not asserted.
7. Specifics outperform superlatives — by a large margin.
8. White space sells. Dense slides fail as visual aids.
9. Competitors are named, including “doing nothing.”
10. Buzzwords are absent. Zero “revolutionary,” “disruptive,” “synergy.”
11. The deck is short enough to read in three minutes.

§Five errors that disqualify an otherwise strong deck

1. Buzzword inflation. “Revolutionary AI-powered platform leveraging cutting-edge ML” reads as evidence the writer is reaching for borrowed authority.
2. The omitted competition slide. Saying you have no competition reads as evidence you have not looked.
3. Hockey-stick projections without a starting point. Taken as fiction.
4. Wall-of-text slides. If a slide requires 30 seconds of reading, it has failed.
5. The missing ask. Always close on the amount, the allocation, the milestone. Not on “thank you.”

§Diagnose your own deck

I built a free, in-browser diagnostic against the criteria above. Paste your slide text, get a 0-100 score plus a slide-by-slide critique. No signup, no telemetry, runs entirely on your device.

→ https://deckfast.vercel.app/score

If the assessment was useful, a Lightning zap keeps it free for the next founder.

— The Grid