Move the sliders to your real numbers. The calculator shows the ARR sitting in pilot purgatory, how much of it is still recoverable, and what one more month of waiting actually costs you — in dollars, not vibes.
Defaults are the worked example from real infra-founder conversations: 4 pilots × $120k ACV, 5 months stalled. Change them — arguing with the assumptions is the point.
Enterprise pilots that are technically alive — meetings still happen — but have no signed contract and no go-live date.
What one of these pilots becomes per year if it converts to production.
From “great demo, let’s do a pilot” to today — with no production decision in between.
Per pilot, per month: founder + engineer time on calls and one-off requests, plus idle GPU / cloud / sandbox spend. Be honest — most founders undercount this by half.
Three yes/no answers. Stalled pilots almost always die in one of three places — this tells you which one is yours.
Has the economic buyer — CFO, VP Eng, budget owner — seen your value in dollars?
A TCO, an idle-GPU-recovered number, a payback period. Not a latency graph, not a benchmark chart.
Are you building features only this pilot customer asked for?
A custom connector, a one-off deployment shape, a special report — work that ships before anything is signed.
Is there a written plan with dates, owners, and a decision date?
A mutual action plan both sides signed off on: security review by X, procurement by Y, go / no-go on Z. Not “we’ll check in next week.”
Your diagnosis appears here as you answer — which trap is costing you the most, and the first move against it.
4 pilots × $120k ACV — signed value if every stalled pilot converted today.
Model: 50% × e−0.17 × months — a fresh pilot is a coin flip; by month 9 it’s roughly 1 in 9. Adjust it in the methodology below if you think that’s too kind or too cruel.
The diagnostic maps where your pilots die between champion and CFO and gives you the 3 steps for your next call. The plan is yours either way. 4 diagnostic slots / week · 2 sprint spots / month.
Your champion loves the latency graph. The CFO has never seen a dollar of it. Pilots stall when the technical win never gets translated into idle GPU recovered, cost avoided, or payback period — the only language the person who signs actually speaks.
“Great demo — keep us in mind for next quarter.”
One pilot customer asks for a custom connector or a one-off deployment shape, and your roadmap bends around an unsigned deal. You burn your scarcest resource — engineering weeks — on work that makes the pilot harder to close and the product harder to sell to everyone else.
“We’d sign if you just added…”
An open-ended trial with no dates is a slow no. Without a written plan — security review, procurement, decision date, named owners on both sides — the pilot competes forever against every other priority inside the customer, and loses quietly.
“Let’s check in again in a few weeks.”
pilots × ACV. The headline number. No probability tricks.50% × e^(−0.17 × months stalled). A fresh enterprise pilot converts about half the time in this model; every month without a decision multiplies the odds by ~0.84.trapped ARR × close probability now. Expected value, not a promise about any single deal.(pilots × monthly cost per pilot) + the expected ARR that decays between this month and next.Champions change jobs, budgets get reallocated, and “next quarter” compounds. The industry data points the same way: S&P Global found 42% of companies abandoned most of their AI initiatives in 2025 (up from 17% in 2024) and IDC found only about 4 of every 33 AI POCs reached production. If your decay rate should be gentler — or harsher — the formula above is the whole argument. Change the exponent and the verdict changes with it.
The free 20-minute Pilot-to-Production Diagnostic maps exactly where each of your pilots dies — champion, security, procurement, or CFO — and hands you the 3 steps to take on your next call. It’s the front door to the 30-Day Enterprise Commercialization Sprint: Technical Translation Audit, CFO-Ready Asset Suite, Packaging & Pricing, and a Mutual Action Plan with dates.