Guide · Orientation
Written for: seed / Series A scientist-founder
A techno-economic analysis is a structured estimate of whether a process can win commercially — what the product costs to make, whether that beats the market or the incumbent, and which assumptions actually decide the answer. At this stage it’s a decision tool, not an academic exercise: an academic cost study optimizes for defensibility to the last decimal; a seed-stage TEA optimizes for changing a decision — go or no-go, this route or that, which experiment to run next.
This guide is written for founders building their first TEA — whether you’re pre-spinout and testing whether there’s a company in the science, or heading into a Series A and putting a model in front of investors. The concept pages stand on their own for any reader; the guide’s judgment calls — every “good enough,” every shortcut — assume it’s your first time through.
Three things a TEA at this stage is not:
And be realistic about fundraising: a TEA almost never wins you a round, though a weak or over-claimed one can lose it. As one climate investor put it — “I’ve never looked at a slide and decided to invest, but I’ve seen a TEA and decided not to.” Build it to do the two things it genuinely can: keep you from getting knocked out, and point your own R&D at what matters.
Every judgment in this handbook — every “good enough,” every 🧭 and 🪜 — is calibrated to one level of maturity: a model substantiated enough, and understood deeply enough by the person driving it, to build conviction — first in you and your team, then in the investors and partners you take it to. Not a FEED study; a model you understand well enough to see where it’s solid, where it’s soft, and what it would take to believe the result yourself.
On the handbook’s TEA Readiness Level ladder — its four stages of build-out, from a one-line process sketch to a diligence-grade model — that anchor sits around Level 2: a first-pass, end-to-end economic picture complete enough to interrogate, reaching up into Level 3 only on the two or three things a sensitivity says actually move the answer. The full ladder is what diligence and a built-out company eventually climb; this handbook calibrates to the rung that earns conviction.
It’s a middle you can miss on either side:
Undershoot it — the toy — and you have one number with no boundary, no sensitivities, no drivers: nothing an investor can interrogate. Overshoot it — the cathedral — and you’ve built a simulation-grade model whose ±30% inputs make its twelve-digit output false precision that cost you three months. Almost everyone’s instinct, especially a scientist’s, is the cathedral. The fastest way to find the anchor: build the whole model thin, end to end, then deepen only the 2–3 things a sensitivity says actually move the answer.
Two kinds of page. The Concepts are a flat set of self-contained definitions — one idea each, neutral and complete: the capital recovery factor is the capital recovery factor whether it’s your first TEA or your fiftieth, so those pages carry no advice. The guide — this orientation, then the five layers — is the opinionated layer on top: it sequences those concepts and says which matter most for a founder’s first model. The rule that keeps them clean: the guide references; the Concepts define.
Throughout the layers you’ll hit two recurring boxes:
Don’t have someone else build you a model — and (tempting as it now is) especially not an AI. The early TEA is never really about the spreadsheet; it’s your own mental model building, and the ability to change it when an investor pushes on a number live. Type every formula, hit your own errors, fix them — that fumbling is the learning!
That’s how to read everything that follows. Start with the five layers: Layer 1 — Structural.