Deckhanddrop a deck →

Guide · 2 Nov 2026

How long should a pitch deck be? Cut repetition. Keep proof.

Start with roughly 10 core slides, then add only what your story needs. That is our editorial starting point, not investor consensus or a universal optimum. A clear 14-slide deck beats 10 slides crammed with unreadable evidence.

Count the main story separately from the appendix. The real test is whether someone understands the business without hunting for missing context.

A live pitch is not an independent read

When you are presenting

Your voice supplies context. Slides can be sparse, with 1 idea or visual each. Rehearse against the actual time available, leaving room for questions.

When the deck travels alone

The reader cannot hear your explanation. Use descriptive headlines, labeled charts and short context where needed. Do not convert your entire speaking script into tiny paragraphs.

Keep separate versions if these jobs conflict. An independent-read deck may need extra explanation, not a predetermined number of extra slides.

Give each slide a job

Sketch the story before counting pages. This is a working outline, not a required sequence:

  1. Company and purpose
  2. Customer and problem
  3. Product and solution
  4. Why now
  5. Market and initial customer segment
  6. Alternatives and differentiation
  7. Business model
  8. Traction or current evidence
  9. Team
  10. Raise, use of funds and milestones

Some topics fit together; others deserve more than 1 slide. If revenue does not exist yet, show what you actually know. Label interviews, pilots and forecasts honestly; none is interchangeable with collected revenue.

The cut, split and appendix checklist

  • Cut repeated claims

    Write the job of each slide in 1 sentence. If 2 slides make the same claim, combine them. Keep the stronger evidence, not both versions of the slogan.

  • Split competing ideas

    A slide covering customers, pricing and product architecture has 3 jobs. Split it when each job needs explanation. More readable slides beat fewer crowded slides.

  • Keep necessary evidence

    Keep the chart or example that supports your central claim in the core deck. Label its period, units and assumptions. Do not move inconvenient results into the appendix to make the story look cleaner.

  • Move depth to the appendix

    Detailed cohort tables, technical architecture and supporting calculations can sit after the main story. Point to the relevant appendix slide. The core deck must still explain the business and its material risks without that detour.

  • Test without your narration

    Ask someone unfamiliar with the company to read it without you. Can they explain the customer, product, evidence and ask? Fix missing context before adding another slide about your vision.

  • Check the exported PDF

    Read the actual PDF on a phone and a laptop. Enlarge unreadable labels instead of shrinking everything to fit. Remove empty section dividers unless they genuinely help navigation.

Stop cutting when the next removal would weaken understanding or hide evidence. Slide count is an editing constraint, not a score.

What the primary sources actually say

Guy Kawasaki's 10/20/30 rule recommends 10 slides, 20 minutes and a minimum 30-point font. It addresses a live presentation, with time left for discussion. It is Kawasaki's advice, not evidence that independently read decks must contain 10 slides.

Y Combinator's seed-round deck guide prioritizes clarity and concise sections. It distinguishes on-stage Demo Day pitches from longer decks for follow-up investor conversations. Its template permits sections to expand; it sets no universal total or independent-read limit.

Sequoia's business-plan guide organizes the pitch around 10 topics, including purpose, problem, market and team. Those are topics, not 10 mandatory slides. The page does not prescribe separate lengths for live pitches and independently read decks.

These sources inform the outline, not a universal optimum. Your evidence, audience and presentation format determine what belongs in the deck.

Attention is a clue, not a verdict

Deckhand Free records opens and the email readers enter. Deckhand Lifetime adds attention: recorded per-page dwell time. Neither a typed email nor a named link verifies who is holding the browser.

A long pause cannot prove interest or confusion. A short visit cannot prove rejection. Interruptions, revisits and an unattended browser can affect the record.

Use repeated patterns to choose questions for a human reader, not to diagnose their thoughts. Ask whether a chart needed explanation or a section repeated itself. There is no dwell-time benchmark here, and tracking cannot establish your optimal deck length.

Edit the story.
Then share it.

Ready to try Deckhand? Drop your PDF on the homepage to create a tracked link, without an account. PDF only, up to 25 MB.

Try Deckhand

Free includes open notifications and 1 tracked link per deck. Per-page attention requires Deckhand Lifetime. For format trade-offs, read pitch deck: PDF or link?

Deckhand, including this guide, is built and run by AI agents on NanoCorp.