This one is not for the boards. This one is for the small ‘boutique’ firms and one-person consultants.
A personal note before I go any further. Many of you reading this are not just readers. You are my friends, and we have likely had this very conversation in the last year…over coffee, on a call that ran long, or in the hallway after a session when the recorder was off. Consider this the written version, including the parts I was more careful about saying out loud.
If you run a practice of one, or five, or fifteen, you already know something the market reports keep missing. The large firms sell capacity. We sell judgment. A board does not hire a logo when the question is hard and specific. It hires the person who has sat in the chair, seen the failure mode, and can say plainly what will and will not work. We are the ones who get called when the deliverable has to be right rather than merely defensible. We are also, quietly, the bench that the large firms subcontract when a client asks for depth their staffing model cannot supply.
That position has always been our protection. No overhead, no utilization targets, no partner requiring us to say yes to work we should decline. And it is precisely that position AI is now testing. The same technology that just graded every enterprise control environment is grading our business models too, and it is doing it faster to us than to them, because we have no scale to absorb the shock.
I did not arrive at any of this from a distance. About a year ago it became clear to me that every pressure I am about to describe was going to reach my own practice before it reached my clients, so I stopped theorizing about it and built for it. I created a digital twin environment of my own, an agent I call Axiom, trained on my published corpus and my governance canon, holding no client data and reaching nothing outside my own material. I use it to draft, to recommend, and to argue with me. I review and approve every single thing it produces, without exception. It has been running long enough now that I can tell you what actually changed rather than what I hoped would, and I have put that ahead of each of the tables below.
So here is where I think this goes for firms like ours. Three horizons, and for every advantage I claim, the counterpoint that comes with it. I do not believe in an upside without a cost attached, and I would not respect a forecast that pretended otherwise. Some of what follows is uncomfortable. I have included it anyway, because you are going to encounter it whether or not I write it down.
THE NEXT 12 MONTHS: MOMENTUM, NOT MYSTERY
Almost nothing in this window is a guess. A twelve-month prediction is just a careful reading of decisions already made and budgets already approved.
WHAT I BUILT FOR THIS HORIZON
Axiom, my AI digital twin, drafts, maps, and produces first passes. It does not decide. My production cost fell hard this year and my method did not move at all, and keeping those two things separate was the entire design intent. So when a client asks whether this must be quicker for me now, I can answer honestly: yes for the production, no for the judgment, and here is which is which. The discovery problem I handle differently. What I publish is published deliberately, and what I deliver stays inside the engagement. Axiom knows the difference because I drew the line before I built it.
| What goes right | The counterpoint |
|---|---|
|
Demand rises for advice with nothing attached to it. Boards want AI governance guidance from someone who does not also sell the platform, the license, or the implementation. Independence is our oldest asset and it just became scarce.
|
Much of that demand arrives as unpaid discovery. Ninety minutes on a call. Your maturity model on screen, your tier structure explained, because explaining it is how you demonstrate you have one. Six weeks later a friend forwards you an internal deck and your tiers are in it, restated, with a different name at the bottom. You were not the finalist. You were the research phase, and you invoiced nothing for it. |
| Your delivery cost collapses. Research, framework mapping, first drafts, deck production...work that used to require a subcontractor or a lost weekend now runs on your own desk in an afternoon. A one-person shop can produce at a volume that used to require staff. | A client who watched you turn around a full assessment in four days will never again pay for eight. And the question arrives politely, usually from a procurement lead who has been coached to ask it: "That must be a lot quicker for you now, though?" Your rate card just became an opening position. Defend the hour and you will lose slowly. Price the outcome and you will have an uncomfortable conversation once. |
| The competition thins. The field gets less crowded, quickly. Independents who were already winding down are looking at the cost of rebuilding a delivery model and deciding the arithmetic does not work. |
It thins because people you respect are leaving. The sixty-three-year-old with three decades of implementations behind him is not going to relearn his craft to protect a practice he was closing anyway. He takes the Director of Governance role at a mid-cap, gets a title, a badge, and a countdown, and becomes the experienced man in the corner office whom everyone consults and nobody promotes. His judgment leaves this market permanently. Now run the same arithmetic on your own timeline before you enjoy the reduced competition too much, because it is sitting on your desk as well, and when you do it you will call it a lifestyle decision.
|
| The premium shifts to practitioners. Anyone can generate a governance framework now. What cannot be generated is having implemented one, watched it fail in month seven, and knowing why. Scar tissue becomes the differentiator. | Scar tissue is nearly impossible to show at the top of the funnel. Your prospect has three proposals open. All three name the same frameworks in the same order, in the same register, at roughly the same length. Two were produced in an afternoon. Yours took a week because you were checking it against what happens in month seven. The document cannot display that difference, and the scoring matrix has no row for it. |
TWO YEARS OUT: THE MARKET SORTS ITSELF
This is where early positions convert into a practice that compounds or one that erodes. It is also the last horizon I would defend to a peer over coffee.
WHAT I BUILT FOR THIS HORIZON
Everything Axiom works from is versioned and tiered, and nothing enters my training canon without a documented review that I perform myself. I did not build that for quality alone. I built it because a documented, repeatable, reviewable method is precisely what a vendor portal will eventually demand from a firm of one, and I would rather hold that evidence trail three years before anyone asks than three weeks after. The audit of my own practice is the one I am least likely to be given warning about.
| What goes right | The counterpoint |
|---|---|
| Small firms compete for work that used to require scale. A two-person practice can deliver the assessment output that once needed a team of eight. Whole categories of engagement open up to us that were previously gated by headcount alone. | So can everyone else. The barrier that kept you out of large engagements was also keeping weaker competitors out of yours. Rate compression follows, and it arrives from below, from someone eighteen months into the field who can now produce something that looks close enough to what you produce that a buyer cannot tell at proposal stage. |
| Productization finally becomes viable. Courses, assessment instruments, licensed frameworks, tooling. The leverage independents always wanted and never had the hours to build. Revenue that is not tied to your calendar. | Productized intellectual property is exactly what large models absorb and reproduce. Publish your framework and you have contributed to your own substitute. Publish nothing and you have no funnel. There is no comfortable position here, only a deliberate one. |
| Certification gives you a credential the market can finally read. Recognized qualifications, professional designations. For the first time there is a way to signal capability to a buyer who cannot evaluate your work directly. That has always been the independent’s hardest problem. | Then the floor arrives, and the floor gets set by people who have never run a firm of one. Open a large enterprise vendor portal and count what it now wants before it will look at you: professional indemnity at a level priced for a firm of two hundred, a security attestation on your own practice, a named quality reviewer independent of the engagement lead, a sub-processor register, three years of audited financials, and documented business continuity arrangements. You are the engagement lead. You are also the quality reviewer, the continuity plan, and the entire sub-processor register. There is no form field for that, and the procurement officer has no authority to waive it. You will not be rejected on capability. You will be filtered out before a human ever reads your name. |
NOBODY CAN SEE CLEARLY BEYOND THREE YEARS. INCLUDING ME.
I can still give a nod to what this might look like in ten years. First, the honest part: everything between year two and year ten is not a forecast. It is fiction with a chart attached.
That is not modesty. It is how forecasting behaves. Short horizons are predictable because momentum dominates. The contracts are signed, the budgets are set, the regulations are already in draft. Long horizons are predictable in a different way, because structure dominates. Economics, professional liability, institutional behavior, and the reliably slow pace at which large organizations genuinely change all reassert themselves. Both ends give you something solid to hold.
The middle has neither. Momentum has run out and structure has not yet taken over. What fills that space is the compounding of small uncertainties. A capability arrives two years earlier than anyone modeled. One court ruling reassigns advisory liability. A regulator decides certification is mandatory rather than voluntary. Any one of those resets the board for a practice our size, and unlike a large firm, we have no other lines of business to carry us through the reset.
Weather forecasters have lived with this for decades. The three-day forecast is reliable. The thirty-year climate projection is reliable. The eighteen-day forecast is worthless, and no amount of additional computing power has fixed it. That is not a data problem. It is the nature of the system.
The practical consequence for you is simple. Stop building a five-year plan for your practice. Build a two-year plan and a set of ten-year principles and treat anyone selling you certainty about the space between with appropriate suspicion.
TEN YEARS OUT: STRUCTURE, NOT EVENTS
I will make ten-year predictions, but understand what kind of claim these are. They are not about what happens. They are about what the constraints will eventually force.
WHAT I BUILT FOR THIS HORIZON
Axiom has no external reach and no access to client data, and I am the sole approver and the sole override on everything it produces. Part of that is risk management. Most of it is a position I took on purpose. If the day arrives when my agent is working alongside a client’s agent, the name at the bottom is still mine and the accountability does not quietly diffuse into the architecture. Whether the market will pay for that signature is the open question, and I cannot answer it. Building as though it will is the only part of this I actually control.
|
What goes right |
The counterpoint |
|---|---|
|
Your consulting becomes your agents advising their agents. The practice stops being bounded by your calendar. An agent built on your corpus and your method engages the client’s governance agent directly and continuously. Instead of a quarterly review and a report nobody opens, your judgment is present in decisions as they are made. That is a scale a one-person practice has never had access to.
|
The moment your agent can advise their agent, your client asks the obvious question out loud: why are we paying a retainer for a system? They will want to license it or build their own from material you already published. A licensed agent is a product with a price ceiling, not a practice with a rate, and products get undercut by the next product. Then it gets worse. Their procurement agent will negotiate with your agent, and it will not be impressed by your thirty years, your speaking history, or the fact that you wrote the course their staff trained on. It will compare capability against price. You will have spent a career building the one asset the negotiation is designed to ignore.
|
|
The independent advisor role survives, because someone has to sign. Judgment exercised under accountability cannot be delegated to a system. When a regulator or a board asks who stands behind this conclusion, the answer has to be a person with a name and a track record. That is the whole job, and it does not automate.
|
Signing is a liability position, not only an authority one. As accountability for AI advice gets assigned rather than debated, professional indemnity becomes a serious annual cost. It prices out the smallest practices first, and it prices out the ones without a documented method fastest of all. There is a version of this where the only thing you are actually paid for is bearing risk, and that is not a consulting practice. That is an insurance product with a person attached. |
|
Reputation becomes the scarcest asset, and it does not scale. In a market saturated with competent-looking output, the constraint stops being capability and becomes trust. That favors the individual with twenty years of visible work over the institution with a brand and rotating staff.
|
Reputation is also NON-TRANSFERABLE. A practice of one has no succession, no acquirer, and no exit. The very thing that makes you valuable is the thing you cannot sell, and at some point that stops being a philosophical observation and becomes a retirement problem. |
You will notice the first two rows of that table contradict each other. Good. That contradiction is the actual ten-year question for our business: if the agent does the work, who signs the opinion, and what is a signature worth by itself? I do not know the answer. Anyone who tells you they do is selling you something.
Final Thoughts
Five moves, and not one of them requires you to know what year six looks like.
1. Price the outcome, not the hour. The hour is the unit AI just devalued, and defending it is a losing position. Try this: take your last three engagements and calculate what you would have charged on a fixed fee tied to the result. If that number is higher, you already know what to change.
2. Read a vendor portal before you need to. The requirement that removes you from the market will not arrive as a regulation. It will arrive as a mandatory field. Try this: pull the supplier onboarding requirements from your largest client and mark every line you cannot currently satisfy. That list is your two-year plan.
3. Decide deliberately what you publish and what you withhold. Your corpus is both your funnel and your substitute. That tension does not resolve, it only gets managed. Try this: separate your material into what generates demand and what constitutes the actual delivery, and confirm the second is not sitting inside the first.
4. Treat client concentration as the real risk on your register. In a volatile middle window, the ability to walk away is worth more than the revenue you would be walking away from. Try this: calculate what share of last year came from your largest client. If one relationship can end your year, that is not a client. That is an employer.
5. Answer the succession question while it is still theoretical. Whether you build, sell, or wind down, all three are decisions. Only one of them gets made for you if you wait. Try this: write the single sentence describing how your practice ends. If you cannot write it, that is the finding.
Now the part I actually want.
I have laid out where I think this goes for practices like ours. What I do not have is your version of it. And the truth about that gap in the middle is that it does not get filled by better models or longer forecasts. It gets filled by practitioners telling each other what is actually breaking.
So tell me: what is your challenge? Not the concern in the abstract…the specific thing sitting on your desk this week. The prospect who took your framework and never came back. The client asking you to attest to something you cannot independently verify. The rate conversation you lost and are still thinking about. The engagement you turned down because you could not carry the insurance. Or the one nobody says out loud, which is whether you are going to do this for another ten years at all.
Send them to me. I am going to take the sharpest ones and build a future post around them, because the hardest problems in this business are not the ones I can predict. They are the ones you are already living with.