How to tell if this is right for you — including how to tell if it isn't.
Why this page exists
Some of this will be obvious. The less-obvious parts are where the mistakes usually happen. The right fit is a person whose own clarity about what matters is already most of the way there — the remaining question is how to move it, not what it is.
If you read this and notice yourself on the wrong side of it, that is useful. Not every fit is a fit, and the fit goes both ways.
Signals
Probably a fit
- There is a part of the work that matters a lot, and you quietly suspect it would move if you got it right.
- You are open to changing how the work actually happens — not only to talking about it.
- You want to understand what is happening inside the work, rather than outsource a black box.
- The work in question has enough context-weight or judgment-weight that generic tooling has not solved it.
Probably not a fit
- You are looking for a course, a prompt pack, or a generic AI overview.
- You want something built without involvement on your end.
- You are looking for inspiration more than implementation.
- The work needs a specialized vendor in a known category — a CRM implementation, a bookkeeping cleanup, a dedicated engineering hire. Those are real problems, but not the one this practice is shaped for.
A few quiet failure modes worth naming
Buying AI education when the real issue is workflow
This is the most common one. A founder takes two courses, bookmarks a dozen threads, subscribes to three newsletters. Six months later the underlying workflow still has not changed. The problem was never that the founder did not know enough about AI. The problem was that the workflow had a shape in which AI had nowhere useful to go. Education without redesign leaves the leverage untouched.
Hiring a technologist to build before the leverage is clear
Build-first engagements produce artifacts that the business does not use. If the leverage has not been isolated — which is the actual diagnostic work — the thing that gets built will almost always miss the right spot by one or two layers. You end up with something sophisticated in a place where sophistication was not the problem.
Treating this as a one-shot project when it is really a way of working
Some engagements are genuinely one-shot. Others are the first instance of a way of working that should keep evolving. Confusing the two means either over-committing to partnership prematurely, or stopping at the first delivery and watching the gains erode. Which one this is becomes legible during Discovery, not before.
Optimizing theatrical AI, not actual leverage
It is easy to spend time on AI moves that look impressive and produce very little. The counter-move is to keep asking: where, in the real work, would a change here actually matter? The leverage question is cheap to ask, expensive to avoid.
Starting with the tool rather than the move
Tools do not name their own use. A model or a tool chosen before the move is named will shape the move toward itself, which is the wrong direction. Move first, tool second.
Why the work is structured the way it is
Discovery is paid because a paid first session filters for seriousness, protects the quality of the diagnosis, and keeps the relationship at equal business stature from the start. A free call tends to become unfocused free consulting, which benefits no one. This is not ideology; it is a choice about what kind of engagement this is.
Applied Leverage is sized after Discovery because the right intervention is visible only once the leverage is located. Pre-packaging every option before the diagnosis is vendor behavior; locating the leverage first and then choosing the intervention is practice behavior. There is a difference.
Partnership is earned, not defaulted to. Most engagements end without becoming partnerships. That is the design.
If you still are not sure
The honest move is to not book a session. Another piece of written thinking elsewhere on the site, or a conversation with someone else in the category, or simply sitting with it for a week, are all reasonable responses to I am not sure.
The first session has a price and a guarantee for exactly this reason: so the decision to book is never ambivalent. Either the leverage emerges and the next step is visible, or I continue at no cost or refund you. That is the only shape in which the decision can be clean.
These are not warnings. They are just what I have seen.