Priced on compute, not value
Charging per token or per seat while the product removes a $90K analyst or a $200K agency retainer. The gap between compute cost and value delivered is almost always $200K–$600K a year.
Pricing against value, tourist churn, inference margin and the buyers you should be selling to: located, priced and ranked in a written report, reviewed and signed by the founder.
7 questions · 3 minutes · no email needed to see your grade
Churn reads as product failure. It's a tourist problem.
Charging per token or per seat while the product removes a $90K analyst or a $200K agency retainer. The gap between compute cost and value delivered is almost always $200K–$600K a year.
High sign-up counts hide two very different customers: committed users who embed the product, and curious "tourists" who tried once and left. When tourists churn at ~43%, the retention number looks like a product problem. It isn't.
AI-native gross margins slide 8–15 points on inference and model spend — quietly turning a SaaS-grade business into a services-grade one. The ~45% AI-native median is the investor floor.
Every tier follows the same path. Only the depth of the diagnosis changes with the price.
Choose the tier that fits your ARR, or take the free grade first and let it point you to one.
2 minutes18 questions on pricing, retention and pipeline. Anything you don't have is flagged in the report, never guessed.
About 15 minutesEvery leak located, priced and ranked, with the fixes in order, reviewed and signed by the founder.
18–24 hours · Starter 48–72Four inputs, one directional estimate from published benchmarks. It is a preview, not the diagnostic.
Directional only. The diagnostic replaces every estimate with a figure from your own numbers.
Are you charging for what it costs you to run, or for the human cost you remove from the customer?
Which of your logos are committed customers vs curious tourists who were always going to leave?
What is your real gross margin after model and inference spend, and is it above the ~45% AI-native median?
Is outreach aimed at the economic owner of the cost you remove, or the technical champion who loves the demo?
Where exactly is NRR leaking — tourist churn, flat expansion, or under-monetised committed customers?
These five are read first. Each tier below states exactly how far the diagnosis goes beyond them.
Pick the tier that matches your ARR. Every diagnostic is reviewed and signed by the founder and delivered in writing.
The 3 most painful AI frameworks on your real numbers — plus the AI governance-gap-to-deal-blocking map. Pricing verdict, cohort split, inference teardown. Sharp first read before the full diagnostic. vs $15K–$50K specialist consultants charge for the AI governance scoping alone.
Seed to Series A. The full board-grade diagnosis — dollar-precise on every framework, priced to your stage. About 0.5% of the ~$675K–$1.35M a sub-$5M-ARR AI startup is typically bleeding.
1% gain ≈ up to $10K/yr at your scale
Seed to Series A. The full board-grade diagnosis — dollar-precise on every framework, priced to your stage. About 0.5% of the ~$675K–$1.35M a sub-$5M-ARR AI startup is typically bleeding.
1% gain ≈ $10K–$30K/yr at your scale
Series A to B. The full diagnosis plus a deeper pass, sized to a scaling AI team. About 0.5% of the ~$1.35M–$2.25M a $5M–$25M-ARR AI startup is typically bleeding.
1% gain ≈ $30K–$150K/yr at your scale
Series B+. Everything in Growth, at the depth a mature AI org needs. About 0.5% of the ~$3M+ a $25M–$50M-ARR AI startup is typically bleeding.
1% gain ≈ $150K–$500K/yr at your scale
A board-grade PDF for each tier. The $450 Starter covers the 3 deepest AI frameworks — all 18 scored including the AI governance readiness map — cover, exec summary, dollar-quantified findings, AI-tourist cohort analysis, inference-margin teardown, and a 30-day priority queue. The Founder tier runs all 18 frameworks at greater depth, with the full 18-framework scorecard, pricing-reset model, and a prioritised 90-day roadmap.
PMP®- and RMP®-certified senior planning engineer (BEng Hons). Every diagnostic is personally run, reviewed and signed by me — no outsourced analysts, no call-booking funnel. The methodology is public: judge the teardowns before you spend a dollar.
Judge the method before you buy. Public teardowns:
If a finding doesn't hold up, a number needs sharpening or an input is missing, we rework the diagnostic at no charge until you can act on it. All sales are final; the rework is the guarantee.
Start with the free grade. When you're ready, the diagnostic puts a dollar figure and a fix on every leak.