The First Thing an Investor Asks For
I've talked to plenty of founders who have great ideas, working products, but are stuck at the same point — no pitch deck ready to present.
Some can do it themselves but it takes 2-3 weeks. Others hire an agency for $1,500-3,000.
The problem isn't just time or money. It's not knowing what investors actually want to see. Sometimes you have all the data but can't tell the story. Sometimes you're a great storyteller but the numbers don't hold up.
AI helps with all of it.
Tools I Actually Use
Primary tools:
- Claude Code +
/pptxskill — generates a .pptx file with structure, copy, and speaker notes ready to go - Claude Code +
/presentation-builderskill — creates an interactive HTML presentation you can present straight from the browser
Alternatives if you don't use Claude Code:
- Gamma.app — type one prompt, get a beautiful slide deck with content instantly. Free tier available.
- Pitch.com — polished templates designed specifically for startup decks, works with AI
- Beautiful.ai — AI auto-layouts your slides. Move one element and the whole design adjusts.
- ChatGPT + PowerPoint — have ChatGPT write the outline + copy, then paste into PowerPoint manually
What a Good Pitch Deck Needs
Most investors want to see 5 things before they agree to a meeting:
- What's the problem — is it big enough? Who has this problem?
- How is your solution different — why you?
- How big is the market — clear TAM/SAM/SOM?
- Are the numbers credible — revenue model, unit economics, projections?
- Is the team capable — do you have the right people for this problem?
Answer all five convincingly and your chances of getting that next meeting go way up.
AI Helps Across the Entire Flow
Step 1: Market Research
AI searches and summarizes market size, competitive landscape, and industry trends automatically.
Market Research prompt:
Do market research for a pitch deck for a [business type] in [country/region].
I need:
1. TAM (Total Addressable Market) — full market size with sources
2. SAM (Serviceable Addressable Market) — the portion we can realistically reach
3. SOM (Serviceable Obtainable Market) — target for the first 3 years
4. Competitor matrix — 3-5 competitors comparing features / pricing / strengths / weaknesses
5. 2-3 industry trends that support the timing of this business
Format the output so it can go directly into slides.
What you get:
- TAM figures with cited sources
- Competitor matrix comparing features and pricing
- Trends that support your timing
Step 2: Financial Model
This is the part that terrifies most founders. But AI can build a model from whatever data you have.
Financial Projection prompt:
Build a 3-year financial projection for a [SaaS / marketplace / retail / service] business.
Assumptions:
- Selling price: [amount] per [unit/month/year]
- Cost per unit (COGS): [amount]
- Customers in month 1: [number]
- Monthly growth rate: [%]
- Fixed monthly expenses: rent [x], salaries [x], marketing [x]
- Starting capital: [amount]
Output needed:
1. Revenue projection years 1-3 (Pessimistic / Base / Optimistic)
2. Break-even point — month [x]
3. Unit economics: LTV and CAC
4. Monthly cash flow — which month does cash run out without funding?
Format as tables ready to copy into slides.
What you get:
- Revenue Projection for 3 years (three scenarios)
- Break-even Point — when do you turn profitable?
- Unit Economics — LTV/CAC ratio
- Cash Flow — will you run out of money before you raise?
Step 3: Storytelling
This is where most pitch decks fall apart — all the data is there but it's told in the wrong order.
AI helps structure the narrative:
"Start with a problem people can relate to → show how big the market is → explain why existing solutions fall short → present your solution → prove it with numbers → ask for funding to do this."
This structure works for investors anywhere in the world.
Full Deck prompt:
Create a pitch deck for [company name / business type] seeking [amount] in funding.
Business info:
- Problem: [describe]
- Solution: [describe]
- Target customer: [who]
- Revenue model: [how you make money]
- Traction so far: [revenue / users / partnerships]
- Team: [size + key backgrounds]
Create these slides: Problem, Solution, Market Size, Business Model,
Traction, Competitive Advantage, Go-to-Market, Financial Projection,
Team, The Ask
For each slide: headline (1 sentence) + 3-4 bullet points + speaker notes.
Step 4: Slide Design and Export
AI generates the slide structure with:
- Layout per page (chart, table, bullet points, visuals)
- Copy for each slide (not walls of text)
- Speaker notes for when you present
Exporting with Claude Code:
For a .pptx file (editable in PowerPoint):
/pptx [paste the outline AI created above]
For an HTML presentation (present from the browser, no software needed):
/presentation-builder [paste the same outline]
Output: 12-15 Slides
| # | Slide | Content |
|---|---|---|
| 1 | Cover | Company name, tagline, date |
| 2 | Problem | The problem you solve + how big it is |
| 3 | Solution | Product/Service overview |
| 4 | Market Size | TAM / SAM / SOM |
| 5 | Product Demo | Screenshots / demo or product mockup |
| 6 | Business Model | Revenue streams, pricing |
| 7 | Traction | Existing metrics (if any) |
| 8 | Competitive Advantage | Differentiation |
| 9 | Go-to-Market | Customer acquisition strategy |
| 10 | Financial Projection | 3 years — Revenue, Expense, Profit |
| 11 | Team | Backgrounds and why you're the right people |
| 12 | The Ask | How much, and what you'll use it for |
Export as PPTX (editable in PowerPoint) or HTML (present from the browser).
Time Comparison
| Task | DIY | With AI |
|---|---|---|
| Market Research | 3-5 days | 2-3 hours |
| Financial Model | 2-3 days | 1 hour |
| Slide Writing | 3-5 days | 2-3 hours |
| Design | 3-7 days | 1-2 hours |
| Total | 2-3 weeks | 1 day |
The Bottom Line
A great pitch deck doesn't come from beautiful PowerPoint slides. It comes from a story that answers every investor question and numbers that hold up to scrutiny.
AI handles both — 10x faster and far cheaper than hiring an agency.
If you have a startup idea in your head but no deck yet — let AI draft the first version. Then refine the parts you know best yourself.





