Signature asset · Free · No gate
Four stages for finding the operating constraint that is actually capping your growth — and only then deciding where AI belongs. Written for founders and operators at 20–500 person B2B software and tech-enabled services companies.
Why this order matters
Stage 01
Take one real week of work for your highest-value workflow — sales-to-onboarding, or ticket-to-resolution, or brief-to-delivery. Write every handoff, every wait, every rework loop. Not the process you documented; the one that happened. Most teams find between four and nine handoffs nobody owns.
Output — A single-page workflow trace with owners and wait times marked.
Stage 02
Attach a cost to each handoff: hours per week, delay in days, or revenue at risk. Precision is not the point — order of magnitude is. Anything you cannot price is a candidate for measurement, not for automation. This is the step that separates an operating problem from an annoyance.
Output — A ranked list of friction points with an estimated weekly cost.
Stage 03
Look for the decision your leadership team makes every week with incomplete information: pricing exceptions, scope changes, hiring, roadmap trade-offs. Repeated decisions made in chat threads are the most reliable source of avoidable operating cost in a growth-stage business.
Output — Two or three repeated decisions rewritten as a standing decision packet.
Stage 04
Only now list candidate AI use cases, and score each against four questions: does it reduce a priced friction point, is the data already reachable, can one person own it, and is it reversible if it fails? Anything that fails two of the four is not a first project, regardless of how good the demo was.
Output — A scored AI opportunity shortlist with a defensible first project.
What the output looks like
Sample structure — real engagement content is confidential.
Sample structure — real engagement content is confidential.
About 90 minutes to read and mark up, and roughly two weeks to run the four exercises with your leadership team. It is written to be used in a working session, not read once.
No. Every step is designed to be run internally by a founder, COO, or head of product. Fascia Labs is useful when you want the read done faster, independently, or with more rigour than an internal team can spare time for.
AI is the last third, deliberately. The first two thirds are about finding the operating constraint. Applying AI before you know the constraint is why most pilots produce nothing measurable.
The Diagnostic runs these four stages with an outside operator, in 7–14 days, with the artifacts produced for you. Start with a free call or score yourself first.