A slide titled 'The elevator to orbit.' over a view of Earth from space, with a cable that rises to a station in orbit.
A Slaide investor deck - space elevator pitch, Ascend theme

A Prompt Structure for Generating an Investor Deck

By · · Updated

This is the prompt skeleton I recommend for an investor deck. My view is that a deck is mostly structure and some judgment, and the structure part is what an AI does well. The judgment part stays with you. So here is the prompt, section by section, and what to fix afterward.

Why write one prompt instead of slide by slide

The usual habit is to open a blank presentation, stare at slide 1, write “Problem,” and get stuck. Structure first and prose second works better. But applying that by hand every time is slow, and it is easy to drift - too much text on slide 4, no clear ask on slide 11.

So write one long prompt instead. Give the AI everything it needs to make the structural decisions, then edit the output. You write the skeleton once and reuse it.

Part one: company context

Founders often skip this part, and it matters most. Paste in:

  • One paragraph on what the company does, in plain words, no jargon
  • The specific problem, with a number attached (“teams lose 6 hours a week to X”)
  • Who pays for it today and what they pay
  • Traction: revenue, users, or pilot customers, with real figures, not “strong growth”
  • The raise amount and what it is for

If you skip the numbers, the AI fills gaps with vague filler like “significant market opportunity.” Garbage in, garbage out applies to decks the same as code.

Part two: audience and tone

Tell it who reads the deck. A seed-stage angel round needs a different deck than a Series A pitch to a fund partner. Specify:

  • Stage of the raise
  • Whether this goes out cold (needs more explanation) or gets presented live (needs fewer words, I talk the rest)
  • Tone: direct, no hype words, numbers over adjectives (yes, you can ask an AI to write plainly)

Ban “revolutionary,” “game-changing,” and “seamless.” They are empty words. A number says more than any of them.

Part three: structure constraints

This part saves the most time. Do not let the AI decide slide count or order freely. Give it the shape:

  1. Cover
  2. Problem
  3. Solution
  4. Product (screenshot or diagram placeholder)
  5. Market size, bottom-up not top-down
  6. Business model
  7. Traction
  8. Go-to-market
  9. Competition, honest positioning not a quadrant chart nobody trusts
  10. Team
  11. Financials, 3-year view
  12. The ask

Twelve slides, not twenty. Say this directly in the prompt: “Twelve slides maximum. If a topic does not fit, cut it, do not compress two topics onto one slide.”

Part four: design constraints

Here is where most AI-generated decks fall apart, and it is worth naming the failure modes so you can prompt around them.

Over-explaining. Left alone, an AI writes six bullet points per slide, each one a full sentence. In my opinion, an audience reads a few words on a slide and then looks at you. So add: “Maximum 5 words per bullet. Maximum 4 bullets per slide. No full sentences.”

Bad chart types. Ask for “market size” and you will get a pie chart. Market size is not proportional data, it is a stacked bar showing TAM, SAM, SOM, or nothing at all. Specify the chart type per slide where a chart appears, rather than trust the model’s instinct.

Wall of text on the problem slide. The model wants to justify the problem with three paragraphs. One paragraph, one number, done. Say this twice in the prompt. A model can ignore an instruction it sees once.

None of this is exotic. It is the same discipline you would apply writing the deck yourself, written down so the AI applies it too.

What the output actually looks like

Ask the AI for a plain text file. This is where Slaide comes in, the deck tool I build. A .slaide file is plain text, and it renders to a web deck, a PDF, and an editable PPTX. Slaide has an Agent Skill and an MCP server, so an AI can write decks directly in that format.

But let me be honest about something here: a heavily hand-designed deck, with custom positioning, layered images and precise typography, may not survive the PPTX round trip cleanly. That is fine for a text-and-chart investor deck. It is a poor fit if you want pixel-perfect brand decks in PowerPoint.

The editor is free for anyone signed in. The only paid part of Slaide is the hosted cloud AI. If you bring your own agent through the MCP server, Slaide itself costs nothing, though your AI provider may charge you. The core is Apache-2.0.

What I always fix by hand

I’d make the same three fixes to any generated deck:

  • The traction slide undersells real numbers or oversells soft ones. Rewrite it with the actual figures.
  • The competition slide is too polite. AI models tend to stay polite about everyone, including competitors. Go back in and make the positioning honest.
  • The ask slide is vague (“raising to accelerate growth”). Replace it with the actual number and three concrete uses of the money.

None of this takes long. But it is the time that separates a deck that reads like every other AI-generated deck from one that reads like a founder who knows their numbers cold.

The prompt, one more time

If you want the short version: give it company context with real numbers, tell it who reads the deck and in what setting, hand it a fixed 12-slide structure, and ban vague adjectives and pie charts. Then read every slide yourself before it goes anywhere near an investor’s inbox. The structure is a starting point, not a finished product.