TEA Handbook

Guide · Layer 5

The Shareable Model

Written for: seed / Series A scientist-founder

When the conversation gets technical, the deck isn’t enough — their engineers want the model. This is a separate artifact from the slides, built for the data room: a model someone can open, follow, and pressure-test without you sitting next to them. Its job is to convey trust that you’ve captured the level of detail the problem requires, and to be genuinely easy to share. Never hand over your real working file with a few cells deleted.

Decide how much model to share

Two versions cover most situations, and which one you reach for depends on the reader:

Either way, build a separate artifact, choreographed for someone who’ll never get into the weeds — anything they can’t follow reads as risk, not rigor:

🧭 Coach’s Read

Think chess, not full disclosure — they feel like they’re exploring, but you’ve already laid the path. Two lines you can’t cross. First, the black box can hide how you hit an efficiency, but it has to declare every real input and output, or the balance around it breaks with no visible cause. Second, abstraction blurs detail — it never shades the answer: banding to the favorable end, or quietly swapping in the best-case capacity factor to slip under $700/t, isn’t protecting IP; it’s a different model wearing the real one’s conclusion. And don’t over-protect — a model no one can sensitize is one no one can trust.

Keep the model clean as you build

You can’t bolt shareability on at the end. The work is far smaller if the internal model has good hygiene throughout: figures referenced rather than hand-typed, sheets organized, the logic traceable. It doesn’t need to be polished — that’s nice, not necessary — but a referenced, organized model converts to a shareable one in an afternoon, where a tangled one takes a week.

(A worked walkthrough, building a model and then turning it into a shareable version, is on the backlog.)

Source the load-bearing numbers

Your model goes to people who can’t rebuild it, so they test it the only way they can: where does each load-bearing number come from? Every figure traces to a source — a vendor quote, a comparable process, a paper, a clearly-stated assumption — and the sources rank: a vendor quote outranks a peer-reviewed value outranks a press-release figure.

🧭 Coach’s Read

The trap that actually surfaces in diligence: source the condition, not just the value. A production credit cited to its headline rate but not to its eligibility rules — the carbon-intensity threshold, the qualifying inputs — survives “what’s the rate?” and collapses on “do you actually qualify?” That second question is the one that gets asked.

Be the one who can defend it

A shared model travels without you, but the moment someone pushes on a figure, the person who built it has to field the question live. That’s why building the model yourself isn’t a detail — it’s the prerequisite for everything in this layer. You can only defend what you genuinely understand.

🪜 Leveling Up — the full diligence-grade data room

The two shareable versions above open a relationship and survive a first technical pass. The deeper artifact is a complete, sourced data room: every assumption documented to its tier of the source hierarchy, the full sensitivity and scenario set, often an NDA-gated un-black-boxed model, sometimes third-party validation of the key figures. Climb to it when the relationship has earned the disclosure — a lead investor in confirmatory diligence, a strategic partner’s engineering team, a project-finance lender — not before. The cost is real: weeks of assembly, a defended provenance for every load-bearing input rather than just the drivers, and exposing detail you’ve deliberately been protecting. At the maturity anchor, the simpler shareable model that keeps its drivers visible is deliberately the right tool — it opens the door the data room later walks through.


That closes the five-layer arc: Layer 1 gave the process its shape, Layer 2 its physical flows, Layer 3 its cost, Layer 4 its drivers — and Layer 5 sends the result into the world honest and intact. The real endgame is the loop Layer 4 opened: a TEA built this way doesn’t just tell an investor “can this win?” — it tells you which experiment to run next.