The 10-minute diligence

Roughly 50 pitch decks land in my inbox every week. If I read each one front-to-back, my evenings are gone and my signal is no better than a coin flip. So over the last year I've ground the first pass down to about ten minutes. Here's what I look at, in order.

Minute 0–2: does the thing exist?

Before I open the deck, I look for something I can touch. A live demo, a staging URL, a video of the product running against real data — any of these beats the best-designed slide. If there is nothing to click, I note that and keep going, but my prior drops considerably.

Minute 2–5: the one metric

I scan the deck for a single number that proves something non-obvious. Retention curves for consumer, a signed LOI for enterprise, an inference-cost delta for infra, a "we ship 3× as many parts per month" for hardware. I am looking for proof that the founder has compressed a loop that used to be slow — not just traction.

If the entire deck's strongest claim is a TAM slide, I close the tab.

Minute 5–8: why this team, why now

I skim the team slide looking for one sentence I couldn't have guessed. "Built X at Y" is table stakes. What I want is: this person spent five years being annoyed by a specific problem, and now the tools exist to fix it. Timing answers are usually either a cost curve moved, a regulation changed, or a model capability crossed a threshold. Anything vaguer than those three is a yellow flag.

Minute 8–10: the reply

By minute eight I know whether I want a call. If yes, I write a two-sentence reply with one sharp question. If no, I still write a two-sentence reply — a real one, with the actual reason. The founders remember, and six months later a "no" turns into a "yes, with this new wedge" more often than you'd think.

---

Most of what gets filtered out here isn't bad work — it's work that hasn't yet earned the 45-minute call. The goal of the ten-minute pass is to find the 2–3 decks a week that deserve one, and to give everyone else a clean, fast answer.