🚀 How to prepare a fundraising deck?

3 tactics to prepare a fundraising deck

👀 Share a teaser deck

• Use a 5-slide deck (Problem, Solution, Market, Traction, and Team) for cold outreach

• Show an obsession with learning and executing fast

  • Include an unexpected insight to pique interest 

  • Show specific wins (e.g. "Grew users 10x in 6 months" vs. "CEO with 10 years experience")

• The teaser deck is solely to trigger curiosity and secure a meeting, not to close the deal

  • Think of the teaser deck as a movie trailer; enough to get people wanting to watch the full movie but not so much to give away everything

  • Investors also get hundreds of pitch decks and don’t have the bandwidth to see everything

🎤 Pitch using a full deck

• For the actual meeting, transition to a 10-slide presentation deck to go deeper

• Start with the same five slides from your teaser deck, and then elaborate on:

  • Business model: How do you actually make money? What are your unit economics? (e.g. If you're a marketplace or e-commerce company, what are your costs of goods sold? What's your revenue split?)

  • Additional context: Whatever else is critical for your specific business and why now? (e.g. If you're in a crowded space and keep getting questions about competitors, make sure your main slides include a clear, compelling insight that shows exactly why you're different and better.) 

  • Anticipated questions: Address frequently asked questions or pushbacks you often receive

• This deck should serve as a visual aid to your storytelling, while avoiding text-heavy slides that distract from your speaking during the meeting

📑 Prepare an appendix

• Build an extensive appendix after your full deck instead of cluttering your main pitch

• Include supplementary information around: 

  • Product and technical architecture 

  • Competition grids and deep dive

  • Detailed traction, unit economics, and metrics (e.g. CAC, payback period, retention curves, activation and engagement etc) 

  • Customer stories and case studies 

  • Risks and mitigation

  • Timeline and roadmap

  • Market sizing assumptions and logic

• Different investors will want to dig deeper into different topics and you don't want to introduce new risk areas unprompted 

• When an investor asks a specific "deep-dive" question, jump to the relevant appendix slide to show you are prepared and data-driven

2 traps to avoid

🚨 Unrealistic projections

• Avoid 5-year financial projections. They are almost always unrealistic or lack ambition

• Instead, provide a 1-2 years milestone roadmap that shows exactly what the investor’s capital will "unlock" and which levers drive your growth

🚨 Premature exit strategies

• At the seed stage, skip the "exit strategy" slide

• Most investors are looking for a 7–10 year partnership and massive, independent winners

• Focusing on an exit too early can signal a lack of long-term conviction or a "quick flip" mentality

1 tool to leverage

📖 Best practice on preparing fundraising materials

• Funding rounds should give 2-3 years of runway

• Leverage tools like Gamma to design and create your fundraising deck

Bonus: 1 trend to spark startup ideas

📈 AI world models will take the spotlight in storytelling

• World models will likely give rise to not just a single game or movie, but an entire new category of generative world experiences

• From fixed scripts to "living" physics: Unlike traditional games that rely on pre-programmed "if-then" logic, world models understand the rules of reality (like gravity, friction, and momentum). This means objects in a story or game will react dynamically to player actions in ways even the developers didn't explicitly script

• The "infinite movie" paradigm: Storytelling will move beyond linear paths into emergent narratives. AI world models can generate consistent 16-bit HDR cinematic environments on the fly, allowing a "viewer" to walk off the set of a movie scene and discover a fully realized, persistent world waiting behind the camera

• Zero-latent interaction: As world models "reason" about the environment, they can compress complex 3D data into latent representations. This allows for near-instant generation of interactive spaces, turning a simple text prompt like "an ancient Roman marketplace" into a navigable 3D experience in seconds

• The death of the "uncanny valley": By training on vast datasets of video and physical simulations, world models can predict natural motion and lighting far better than current generative video. This leads to characters and environments that behave with a high degree of "physical common sense," making virtual worlds feel indistinguishable from reality