Guide

How to build a pitch deck investors without a deep tech background can understand

Most deep tech decks read like technical specifications. The fix is not simpler language but a different order: problem and market first, mechanism second.

Most investors in the room are generalists on your specific mechanism, including at specialist funds. They are evaluating market, team and business model at least as hard as they are evaluating the science. A deck written as a technical specification asks them to do the translation themselves, and most will not.

What is a pitch deck?

A pitch deck is a short visual narrative that helps investors decide whether a company merits further investigation and whether its risks are worth underwriting. It connects problem, market, solution, team, business model and evidence into an argument an investor can understand and repeat. Investors use that argument to champion the deal internally. Deep tech investors expect founders to state the core risk plainly rather than pretend technical evidence has removed every uncertainty.

Deep tech founders write decks like specifications because that is who they normally explain their work to: engineers, researchers and industry peers. A specification documents how a product works and the evidence supporting its performance. A deck has to explain why the problem deserves attention, why this company can solve it, and why the business could become valuable.

Dense terminology shifts the work of interpretation onto the investor. A technically accurate deck still fails when a generalist partner cannot explain the company in a partner meeting.

Why should the problem come before the mechanism?

Because investors need a frame for every technical claim that follows. A founder who opens with material science, system architecture or a proprietary process is asking the audience to judge a solution before they know what it solves. Name the affected buyer and the costly limitation that buyer faces.

Good problem framing connects industry scale to individual impact. A battery startup might quantify capacity lost across grid storage projects, then show how that degradation changes the economics of one operator's installation. The first supports the market opportunity; the second makes the reason to buy concrete.

Your market slide should build on that customer-level problem. Estimate how many buyers have it and what they already spend managing it. Investors can then trace market size back to purchasing behaviour rather than accepting a category estimate with no connection to revenue.

Place "why now" before the technical explanation too. Deepchecks recommends explaining the combination of recent technical progress and market willingness that makes the business possible now. A lower component cost may make deployment economical; a new customer requirement may create demand. Name the specific change.

What should the "how it works" slide actually prove?

That the product is hard to build and hard to copy. Investors rarely need a lesson in the underlying science during the main pitch. They need to know which technical achievement creates an advantage, what makes it difficult to reach, and how long it lasts.

Use a visual metaphor to make the mechanism visible and a precise analogy to explain why it is difficult to reproduce. Cobalt Communications puts the distinction as well as anyone: a metaphor shows, and an analogy explains. An image of a lock and key can represent molecular selectivity. The analogy alongside it should explain what makes your key unusually precise and why making another requires knowledge competitors lack.

Open the slide with the advantage in business terms. "Our sensor identifies contamination at one tenth the concentration required by current systems" tells an investor what the mechanism achieves. The visual can then carry the scientific principle, with a small number of annotations connecting it to the strongest evidence you have.

Technical difficulty alone is not defensibility. A competitor may reproduce a hard mechanism once your product proves the market. Identify the barrier that survives a competitor understanding your approach. Intellectual property and proprietary data are the usual candidates; customer integration and exclusive distribution create another when replacing you would impose real cost or delay.

Make each barrier causal and specific. A data moat matters when exclusive operating data measurably improves calibration, prediction or yield. A patent matters when its claims cover the commercially useful implementation rather than a narrow laboratory method. Lock-in matters when contracts, certifications or embedded workflows slow a competitor's adoption.

A loose analogy destroys the credibility it was meant to build. Ask a technical colleague to find where the comparison stops matching the science, and revise any image implying a capability the product does not have.

What belongs in the core deck, and what belongs in the appendix?

Keep a slide in the core deck only when an investor needs its message to follow the investment case on a first read. Move material to the appendix when it verifies a claim without advancing the problem, opportunity, solution, defensibility or business case.

Give each core slide one claim and the clearest chart, diagram or image that supports it. Duygu Dülger, who builds deep tech decks for a living, puts full datasets, detailed methodologies, architecture diagrams and additional validation in a technical appendix. Where a performance claim rests on several measurements, the core slide carries the comparison that affects the decision and the appendix preserves the conditions and calculations.

Treat the appendix as a diligence resource, not the back half of an oversised deck. Organise it around what a technical reader will ask: how you collected the data, which assumptions shape the model, where the current evidence stops.

Founders overload the core deck because they fear simplicity reads as unrigorous. Dense slides make rigour harder to assess, not easier, because nobody can tell which evidence supports which claim. Rigour comes from traceability.

Test every borderline slide by removing it and reading the deck again. If the argument breaks, it belongs in the core. If only the proof disappears, it belongs in the appendix.

How should the narrative change for each kind of investor?

One investment thesis should support several versions of the deck, because each investor tests a different part of the company. Define the customer problem, opportunity, mechanism, defensibility and primary risk once, then adjust depth and evidence without changing the claims.

A generalist partner is looking for familiar business patterns and a reason to champion the deal internally. Keep the mechanism visual, tie it to a customer outcome, and state the primary risk in plain language: manufacturing yield must stay above a given threshold for margins to work. Do not open with the material science behind that threshold.

A technical associate will stress-test assumptions, benchmarks and causal claims. Give them an annotated mechanism diagram, comparison data and direct access to the appendix. A specialist fund may pull that detail into the first meeting because its investors already know the category, but even there the depth should help them judge defensibility and commercial viability rather than reconstruct a research paper.

Every version names the same underwriting risk. The generalist needs its business effect, the technical reader needs evidence testing the assumptions behind it, and the specialist needs enough to judge whether it can be reduced through engineering, IP, manufacturing progress or customer adoption.

What do Cloudphysician and Sevenloop show about translation?

Both are methodology proof, not pitch deck case studies, and worth reading as such.

Cloudphysician works in healthcare AI, where precise patient-vitals data has to stay accurate and stay legible to non-specialists making high-stakes decisions. We made the technical value readable without assuming every reader shared the same product knowledge.

Sevenloop had the related problem in metal manufacturing: one narrative that stayed credible with technical readers and understandable to non-technical decision-makers. Each audience got an appropriate entry point without the company ending up with two conflicting product stories.

The order is the transferable part. We diagnose the product and the audience's existing knowledge first, adapt the narrative second, and decide how visuals carry the explanation last. In one client engagement we replaced a dense seventeen-minute film with five audience-specific videos totalling about nine minutes, one of them investor-facing. A deck works the same way: a stable core narrative, audience-specific slide choices, and an appendix that enters when needed.

A deck benefits from that order too. Understand the mechanism and the defensibility well enough to decide what each investor must learn, then choose the analogy, the depth and the visual structure.

How do you restructure a deck you already have?

Diagnose the product before opening a slide template. Write down the customer problem, the technical advance, and why a capable competitor cannot reproduce it quickly. Back defensibility with specifics: protected IP, proprietary data, manufacturing knowledge, customer lock-in. If you cannot explain the advantage plainly, more diagrams will not save it.

Build a narrative for each likely reader. A generalist needs market and core risk before mechanism. A technical associate needs enough detail to test your claims. A specialist may expect deeper comparison with competing approaches. Same central argument, different depth.

Turn the narrative into slides only once the argument works. One claim per slide, written as the headline. Choose the visual that proves it. Everything else goes to the appendix.

Pressure-test before sending. Show each slide to a smart person outside your field and ask them to state its claim after five seconds. Rewrite anything they misread. Split anything carrying two ideas.

State the largest unresolved risk directly. Name the assumption, explain how you are testing it, and say what evidence would reduce the uncertainty. Investors will find it in diligence anyway, and a clear account arms your internal champion.

What is the deck actually for?

To make the investment understandable enough that an investor can explain it without you in the room. Present the customer problem, the business opportunity, the technical advantage and the primary risk clearly enough that they can assess the case and defend it to their partners.

The deck's job is not to prove you are smart. It is to make the investor feel smart enough to say yes.

FAQ

Should a deep tech deck be shorter or longer than a software deck? The core deck should be about the same length. The difference is the appendix, which carries the datasets, methodology and validation a software deck rarely needs.

How technical should the first meeting be? Set it by the reader, not by your comfort. A generalist partner needs the mechanism visual and the risk in plain language. A specialist fund may want comparison data in the first conversation.

Is a pitch deck a design problem or a positioning problem? Usually positioning. If the team cannot agree on what the company does, the deck will show it. Our positioning alignment check is a faster way to find out than a redesign.

What if the technology genuinely cannot be simplified? Then simplify the claim rather than the science. State what the mechanism achieves in business terms, and let the appendix carry the proof.

Book a 30-minute audit →