AI Stack Economics — enterprise AI value management

How to use this page

What this is

One question, answered in layers

What does AI really cost your company — and what is it worth? The page starts with your WHOLE company (the Enterprise view: total AI spend, total value, ROI). From there you zoom into one team, then into one use case. Every number recomputes the moment you type. Nothing leaves your browser until you hit Save.

What to do

Three steps, five minutes

·1 · Type your company's name in "Model a company" — the calculator fills itself in from public filings. Private company? Answer three quick questions instead. Or skip it and use the built-in example.
·2 · Read the Enterprise view: is the total spend the right size? Which teams create the most value? Click any team's "Open" to zoom into its program.
·3 · Fix what's wrong. Every screen pairs its numbers with the assumptions behind them — change anything (seats, usage, prices, value) and watch the answer move.

What you get

Numbers you can take to a board

·An enterprise AI P&L — spend, value, and ROI by team (CFOs report this to investors now)
·Use-case economics: total cost, ROI, payback, cost per interaction
·Strategy studio: what a 20% budget cut removes, what a provider outage or brownout costs, which teams earn their spend
·A written executive narrative you can copy straight into a deck
·Save, reopen, and compare scenarios side by side

The honest part

Modeled until you make it yours

Every figure starts as a model — built from your assumptions and public list prices, not your invoices — and the page says "modeled" wherever that's true. That's the point: it hands you a defensible starting structure, and every single input stays editable. The more real numbers you put in, the more the answer belongs to you.

Model a company

Type a company name or ticker — the calculator pre-populates from its public filings (revenue → archetype, sector → vertical, headcount → modeled users) as a starting point you can edit.

Modeled enterprise program: $15.5M/yr AI spend across 8,562 seats (26% of the workforce) creating $80.1M/yr in modeled value — 369% ROI on year-one TCO. Operations leads on modeled net value.

Modeled, not observed — the whole view extrapolates this scenario's per-seat economics across 33,333 employees. Edit any function's adoption or intensity below; drill in to model one function's program in full.

AI P&L by function

Each row is that org's modeled program: seats = headcount × adoption; cost = its usage + a seat-share of the shared platform; value = the productivity lever at its scale. Adoption and intensity are editable modeled defaults.

Function	Headcount	Adoption	Intensity	Seats	Cost / yr	Value / yr	Net / yr	ROI
Operations COO	18,333	Operations adoption percent%	Operations usage intensity multiplier×	4,583	$5.78M	$42.35M	$36.57M	632%
Sales CRO	3,000	Sales adoption percent%	Sales usage intensity multiplier×	1,350	$2.56M	$12.47M	$9.91M	386%
Technology CIO / CTO	2,000	Technology adoption percent%	Technology usage intensity multiplier×	1,300	$4.13M	$12.01M	$7.88M	191%
Marketing CMO	833	Marketing adoption percent%	Marketing usage intensity multiplier×	417	$880K	$3.85M	$2.97M	337%
Corporate / CEO office CEO	667	Corporate / CEO office adoption percent%	Corporate / CEO office usage intensity multiplier×	334	$776K	$3.08M	$2.31M	297%
Finance CFO	500	Finance adoption percent%	Finance usage intensity multiplier×	275	$756K	$2.54M	$1.78M	236%
People / HR CHRO	467	People / HR adoption percent%	People / HR usage intensity multiplier×	210	$355K	$1.94M	$1.59M	447%
Legal & compliance GC / CCO	233	Legal & compliance adoption percent%	Legal & compliance usage intensity multiplier×	93	$237K	$861K	$624K	264%

EvidenceEModeledheadcount splitEModeledadoption, intensity & value extrapolation — see Method & sources for the grading scale

· No workforce split on this scenario — modeled a 33,333-employee enterprise at typical functional shares (type your company in the quick start, or edit headcounts, to replace it).
· Per-function value is the productivity lever only (seats × hours saved × loaded rate); the configured use case's structural levers (process cost, revenue, risk, cycle time, retired tech) are counted once at enterprise level.
· Shared platform, governance, and FinOps program costs are counted once and allocated by seat share; each function adds only its variable usage cost.