Concept

TEA in a pitch deck

Overview

A TEA in a pitch deck is the highly compressed representation of a techno-economic analysis for a financing audience — typically a headline number, a benchmark against the incumbent, the few drivers, and a one-line basis, on one or a few slides. It is the most lossy form a TEA takes, built to orient and persuade rather than to document.

Body

What it is. The deck representation reduces a multi-sheet model to a slide’s worth of conclusion: the headline number with its band, a comparison to the incumbent on the metric that matters, the handful of drivers, a scenario range, and a one-line statement of basis. It selects and compresses; it does not derive. The full analysis sits behind it, reached only if the reader chooses to look closer.

Its job is orientation, not documentation. Unlike the shareable model — a workbook a diligence team interrogates — the deck is read quickly by a non-specialist deciding whether the opportunity is worth a closer look. So it carries the conclusion and the reason to believe it, and routes the detail to the model behind it. The deck persuades; the model defends. The two are different artifacts for different moments, and a deck is not a substitute for the model it points to.

Honest compression keeps the qualifiers. Even at one slide, the figure has to keep its band, its basis (boundary, capacity factor, gross or net), and an honest label on whether it is shown net of credits. The compression is in detail, not in honesty: a deck legitimately drops the derivation, the line items, and the secondary metrics, but the qualifiers that make the headline mean something are not detail — they are part of the number.

It amplifies every communication failure. Because so much is cut and the slide may be all the audience ever sees, the failure modes of the other communication concepts — false precision, an unstated basis, a cherry-picked headline, a credit-flattered net figure — are at their most consequential here. The deck is where a true-but-misleading number does the most damage, simply because it reaches the most readers with the least context.

Limits & typical error

• Maximum lossiness meets maximum reach. The deck shows the least and is seen the most, so an omitted basis or band misleads more readers than the same omission buried in a model would; the slide’s reach magnifies whatever it leaves out.
• Persuasion pressure pulls selection toward the favorable. The deck’s purpose — winning interest — biases the choice toward the best-case scenario, the net-of-credits headline, and the one comparator the process beats. A true-but-cherry-picked slide is the characteristic failure, and it contains no false number.
• A number with no band reads as a promise. On a slide, a bare “$800/t” with no ± and no basis is taken as a committed figure rather than a ±30% estimate, setting expectations the analysis cannot honor — an acute case of false precision created by compression.
• The benchmark can mislead on basis. A comparison to the incumbent on mismatched boundaries, or across different maturities (a first-of-a-kind plant against a decades-optimized one), overstates competitiveness; the deck rarely has room to state the comparison’s basis, which is exactly when that basis is most often dropped.
• A figure that points to nothing collapses on contact. A deck number with no defensible model behind it — no provenance a follow-up question can reach — is a claim that fails at the first diligence query; the persuasive slide is only as sound as the analysis it routes to.

See also

• The headline number — the central figure a deck is built around, which must keep its band and basis even when compressed.
• Benchmarking against incumbents — the comparison a deck uses to show competitiveness, and the basis trap that comes with it.
• Driver / key parameter — the few inputs a deck cites as the reason to believe the headline.
• Scenario (assumption bundle) — the case range a deck draws on, and the temptation to show only the favorable one.
• The shareable model — the diligence-oriented artifact a deck routes to, distinct from the deck itself.
• False precision — the failure most amplified by the deck’s compression.
• Provenance & defensibility — what must stand behind a slide figure for it to survive the first question.

Mini-example

Green ammonia on one slide, compressed honestly: “Green NH₃ ~$800/t (±~30%, power-to-ammonia, CF ~0.90) vs gray ~$200–400/t today; gap closes to $260/t net of the clean-H₂ credit. Drivers: power price ($40/MWh) and electrolyzer efficiency.” That single line carries the headline with its band and basis, the benchmark, the gross-versus-net distinction, and the two drivers — and it points to the model where each is sourced.

Separately, to show the same slide turned misleading: rendered as “Green ammonia: $260/t — cost-competitive today,” it drops the band, the basis, the ~$800/t gross figure, and the credit-dependence. Each omission reads as “just simplifying for the slide,” but together they promise a competitiveness the ±30%, policy-dependent analysis does not support — the deck’s compression weaponizing four small omissions at once.

See also

• The headline number
• Benchmarking against incumbents
• Driver / key parameter
• Scenario (assumption bundle)
• The shareable model (trade-secret-safe)
• False precision
• Provenance & defensibility