Most independent experts sell the engagement well and design it badly. The design is what decides whether the client renews — and whether you still want them to.
A dispatch on the part nobody writes about: what a paid mentorship engagement is actually made of, once the selling is over.
TL;DR — the answer first: a mentorship engagement that works has five decisions baked into it before the first session. A ninety-day term, because it is the shortest window that can carry a real result. A fixed cadence that never moves. A session shape organized around decisions rather than topics. A deliverable that is a running system, not a document about one. And bounded access between sessions, stated out loud, so generosity has an edge on it. Get those five right and the engagement mostly runs itself. Leave them undefined — which is what most experienced people do, because the conversation flows naturally and they assume the shape will emerge — and you get the familiar outcome: pleasant sessions, no visible change in the client's business, and a quiet non-renewal at day sixty.
It looks like this. Three months in, the calls are good. You like the person. They arrive with interesting problems and leave sounding clearer. They tell you the sessions are valuable and they mean it.
And nothing in their business has changed.
When the term ends, they say something warm and vague about circling back after the quarter. They do not circle back. You replay the calls trying to find the moment it went wrong and cannot find one, because there wasn't one. The engagement was well-delivered and structurally incapable of producing a result. Every hour was spent on discussion, because no hour was designed to end in a commitment.
This is not a competence problem. It is almost always a design problem, and it is most common among the most experienced people, for a reason that is easy to miss: when you genuinely know a domain, conversation about it is effortless and feels productive. You can fill sixty minutes with real insight without ever asking the person to do anything. The expertise itself is what hides the missing structure.
The structure below assumes a particular shape of person, so it is worth checking whether it is yours before you spend the time.
You have somewhere between fifteen and thirty years inside one operating domain — running plants, running clinics, running a book of business, running engineering teams, running a trade. Not a general management career; a specific field you can still get your hands into. You have one or two people who have already asked you to advise them, formally or otherwise. Your income today comes from a salary, or from project work you charge hourly, or from fractional arrangements that were shaped by whoever hired you rather than by you. On AI you are somewhere between "I use it for email" and "I have built a couple of workflows" — not an engineer, and not a sceptic either. And the thing actually blocking you is not finding a client, and not what to charge; it is that you have never seen the inside of a well-built engagement, so you do not know what you are supposed to be constructing.
If that is close, the rest of this is the blueprint. If your bottleneck is earlier — nobody has said yes yet — The First Call is the dispatch that comes before this one.
Term length is the first thing to fix, because everything else inherits from it.
Short engagements — a session, a week, a month — cannot carry a result. You diagnose accurately, you hand over a recommendation, and you leave before anyone discovers whether it survives contact with the business. That is advice. Advice is a legitimate product, but it is priced and sold differently, and it is not what someone means when they ask you to mentor them.
Open-ended engagements fail in the opposite direction. With no end date, there is no moment where the work is assessed, so it degrades into a standing conversation that both sides are slightly relieved to be having and neither can end without awkwardness. The client stops bringing hard problems. You stop pushing. It runs for another year at a rate neither of you would agree to today.
Ninety days sits between the two failure modes. It is roughly twelve working sessions — enough to install one thing, watch it run for a few weeks, and fix it once when reality disagrees with the plan. It has a natural end, which means it has a natural renewal, which means the question "is this working?" gets asked on a schedule instead of being avoided indefinitely.
One rule: pick one outcome for the ninety days, not three. A term with three objectives has none.
Weekly at first, or every other week once the between-session work gets heavy. The frequency is the small decision. The fixed slot is the large one.
A standing Tuesday at 9 that never gets rescheduled does something a more generous but floating schedule cannot: it puts a recurring deadline into someone else's week. A meaningful part of what a mentee is buying — often the largest part, though few will say so — is the existence of a person who will ask on Tuesday whether the thing got done. Move the slot twice and the mechanism is gone. The engagement becomes optional, and optional things lose to urgent things every week for the rest of the term.
Which also means you should be careful about how much you offer. Weekly for ninety days is twelve sessions you are committing to hold at the same hour regardless of what your own quarter looks like. If you cannot hold it, offer fortnightly and hold that.
Here is the session shape that separates engagements which produce change from engagements which produce clarity:
The test is simple. If the session ends without a named commitment and a date, it was a conversation. Conversations are enjoyable and they do not renew.
This is the part the book is most direct about, and it is worth quoting exactly, because the contrast is the whole argument. From Chapter Eight of The Mentor Economy, describing two businesses solving the same problem — one hiring a large consulting firm, one bringing in an independent operator:
"The Founder does not produce a slide deck. They sit down with the owner, walk through their lead pipeline, and within two weeks have built a new intake system that runs on AI. Within six weeks the new system is producing thirty percent more qualified leads. Within six months the business has a documented playbook the team actually uses, an AI-powered system that runs continuously, and a Founder on retainer who answers the phone when something needs adjusting."
That passage is the book's illustration rather than a case study — it opens with "imagine two small businesses" — so read the numbers as the shape of the argument, not as a measured outcome you should expect. But the structural point holds regardless of the figures: the firm leaves a binder, and the owner is left with a team that does not know how to execute it. The independent operator leaves something that is running.
Applied to your engagement: by day ninety, something specific should be operating in the client's business that was not operating on day one. An intake process. A pricing model they can apply without you. A weekly review that happens whether or not you are in the room. A hiring rubric. Write it down — but the document is the receipt, not the product.
This is also the cleanest reason experienced operators can charge what they charge, and the mechanism behind selling judgment rather than time. You are not selling twelve hours. You are selling the thing that is still running in month seven.
Between-session access is where solo practices bleed. The instinct of a generous, senior person is to say "reach out any time" — and to mean it. Then a client who is genuinely in trouble reaches out on a Friday night, you answer because you would rather help than not, and a norm has been set that you will resent within a month and cannot renegotiate without seeming to withdraw something.
The fix is not to be less available. It is to give the availability an edge:
"Message me any time. I answer once a day on weekdays. If something is genuinely on fire, say so in the message and I will move."
Say it in the first session, before there is a situation. It is generous, it is honest, you can sustain it for years, and it delivers the thing the client actually wanted — which was never instant response, but the certainty that they would not be stuck for a week. This is the same leak, caught earlier, that The Scope You Never Wrote Down describes in its later, more expensive form.
Since the design is now fixed, the leverage question becomes easy to answer.
AI belongs in the surfaces around the session: the summary and commitment list sent within the hour, the first draft of the playbook you are building together, the recap that lets a client find what you said in week three without your help. Those are the parts that are real work, that make you look organized, and that you will silently stop doing by week six if they are manual.
AI does not belong in the thirty-five minutes. The live decision is the entire reason the person hired a human with scar tissue instead of subscribing to something. Automate the paperwork around the judgment; do not automate the judgment.
Structure feels like a downgrade when you are good at your field. A fixed agenda and a written commitment can seem procedural next to the freewheeling conversation you know you are capable of, and there is a quiet worry that imposing a shape will make you look junior — like someone compensating with process for what they lack in judgment.
It reads the other way to the person paying. An engagement with a shape signals that you have done this before and know how it goes wrong. And the structure is not there to constrain your expertise; it is there to make sure your expertise lands somewhere other than the client's short-term memory.
You already have the expensive part. Ninety days, a slot that never moves, decisions instead of topics, a system instead of a binder, and access with an edge on it — that is the container it goes into.
The chapter behind this dispatch — why an independent operator beats a consulting firm in the layer of the economy where most businesses actually live, and what the two engagements look like side by side — is Chapter Eight of The Mentor Economy. Get your copy →
Ninety days is the shortest term that can carry a real result, and long enough that both sides stop performing. Anything under about six weeks turns into advice — you diagnose, you hand over a recommendation, and you leave before anyone finds out whether it worked. Anything open-ended tends to drift into a standing conversation that neither side can end gracefully. Ninety days gives you roughly twelve working sessions, which is enough to install one system, watch it run, and fix it once. Renew in ninety-day terms after that if the work continues.
Weekly or every other week, at a fixed time, for a fixed length. The cadence matters more than the frequency: a standing Tuesday at 9 that never moves outperforms a more generous schedule that gets rearranged. What you are selling is the existence of a deadline in someone else's week. Fortnightly is usually right when the work between sessions is heavy; weekly is right early on, when the person is still learning what to bring you.
Three things, in this order: what happened since last time, the one decision in front of them now, and what they will have done before the next session. Most people organize the hour around topics instead of decisions, which is why sessions feel useful and produce nothing. If the session ends without a named commitment and a date, it was a conversation, not a working session.
A working system the person keeps, not a document about a system. In practice that means the thing you built together is running in their business when you leave — the intake process, the hiring rubric, the pricing model, the weekly review. Write it down, certainly, but the written artifact is the receipt, not the product. A binder of recommendations is what a consulting firm leaves behind, and it is the reason those engagements so often end with nothing changed.
Offer bounded access, not unlimited. Something like "message me any time; I answer once a day on weekdays" is generous, deliverable indefinitely, and does not require you to be reachable at 11pm to honor it. Unlimited access is the single most common source of uncompensated scope in solo practices, and it is rarely what the client actually wanted — what they wanted was certainty that they would not be stuck for a week.
Build the whole thing around a single fixed weekly slot outside your working hours and keep the term short. One mentee, one recurring session, a ninety-day term, and asynchronous review between sessions. The constraint is real, so put it in the design rather than absorbing it privately — a structure that fits four hours a week and holds is worth more than an ambitious one you quietly stop honoring in week five.
As soon as the engagement has a shape. Hourly is the right instrument for a single diagnostic conversation and the wrong one for a ninety-day build, because it prices the meeting rather than the result and it punishes you for getting faster. Once you can describe what the ninety days installs, price the term.
Author of The Mentor Economy and co-founder of MentorMe. He writes about turning hard-won expertise into AI-leveraged one-person businesses.
The Mentor Economy is the full system — free, you just cover $9.95 shipping.
Get Your Free Copy