Essay · Agentic Wealth
The 100-client ceiling is a software problem
Every good advisory practice eventually stops growing at about the same size. The industry decided that number was a fact about advisors. Most of it is a fact about software — and the software just changed.
The number every practice finds
Every good advisor eventually runs into the same number. Somewhere near a hundred clients — for some it's eighty, for some a hundred and twenty — the practice stops growing. Not because the advisor stopped being good at the work, and not because demand dried up; the referral list is often longest at exactly the moment a practice caps itself. It stops growing because the week is full.
The industry has largely made peace with this. We gave the cap flattering names: a lifestyle practice, staying boutique, keeping it personal. Some of that is real — some advisors chose the shape on purpose, and nothing here argues with them. For most, though, the cap wasn't a choice. It was a wall, and the names came afterward, the way names do.
The stakes are larger than any one practice. McKinsey projects the industry will be short roughly 100,000 advisors by 2034 relative to the households that will want one. I've written about that shortage before as a supply problem — a profession that has not figured out how to produce enough of itself. This is the other half of it. A gap that size does not close by recruiting harder, and every capped practice is capacity the industry already trained and cannot use. So the question is worth more than curiosity: is that number a fact about human beings, or a fact about software? The industry has treated it as the former for forty years. Most of it is the latter — which means the shortage is, in part, a software problem too.
What the ceiling is made of
Look at how the capped advisor's week is actually spent. The research on advisor time use keeps landing in the same place: something like a fifth to a quarter of the working week goes to clients — call it eight or nine hours of actual conversation. The rest goes to everything wrapped around those hours: preparing for the meeting, documenting it, the follow-up it generates, account paperwork, transfers and signatures, rebalancing, tax-loss review, compliance documentation, reconciling what the custodian says against what the performance system says.
And underneath all of it, the part nobody puts on a slide: getting a household's information into usable shape in the first place. The spreadsheet where a client keeps their real estate. PDF statements from the old 401(k). The trust document, the K-1s, the held-away accounts that exist only as a login. Somebody has to read all of it, key it in, keep it current, and notice when it changes. That is work nearly no human actually wants. It is also the work most often skipped by people building software for this industry, because it is unglamorous and it does not demo well — and it is one of the largest things standing between a good advisor and a second hundred families.
Almost none of that is advice. It's the packaging around advice. And nearly all of it shares three properties: it is testable, repeatable, and settled. There is a right answer, the right answer is checkable, and no one's judgment is exercised by producing it by hand for the ninetieth time.
The wrong ceiling
There is a real human ceiling in this business. People can hold only so many genuine relationships — the anthropologists put stable social relationships at roughly a hundred and fifty, and advice-grade relationships are more demanding than social ones. If advisors were spending forty hours a week in deep conversation with families and topping out at a hundred relationships, the ceiling would be human, and software would have nothing to offer it.
That is not what is happening. The advisor at the wall is spending eight or nine hours a week in relationships, not forty. The relationship capacity is barely drawn on. What is exhausted is the administrative surface area, which grows with every account, every household, every held-away asset, every regulation layered on since the eighties. Advisors are not hitting the relationship ceiling. They are hitting the administrative ceiling, several floors below it.
The relationship ceiling is also softer than it looks. Part of what limits how many families a person can carry well is memory — what the daughter's diagnosis was, why they sold the rental in 2019, which promise you made in March that they are still counting on. Advisors compensate with notes nobody has time to re-read, and the quality of a relationship quietly becomes a function of what you happened to recall. Intelligence built to hold that well — to retain a household accurately and hand it back at the moment it matters — doesn't substitute for the relationship. It extends how many a single advisor can carry without any of them getting thinner. That is a different claim than efficiency, and it may be the more important one: software that raises the human ceiling, not just clears the administrative floor beneath it.
A software problem, finally solvable with software
That is the sense in which the ceiling is a software problem: the constraint that actually binds is made of exactly the work software can now carry. The line I've drawn elsewhere applies here without modification — work that is testable, repeatable, and settled belongs on rails; work that requires knowing a family, weighing a trade-off, or sitting with uncertainty belongs to the advisor, and always will. The hundred-client ceiling is what happens when the first kind of work is done by hand and bills itself against the second kind's calendar.
Every prior generation of advisor technology promised leverage and delivered a new inbox: a portfolio system that needed reconciling, a CRM that needed feeding, a planning tool whose assumptions needed babysitting. Useful, each of them — and each another thing the advisor operates.
Agents are different in kind, not degree. They do the work against a written methodology, show what they did, and improve with use. The raw material a household arrives with — statements, spreadsheets, PDFs, the accounts held somewhere else — gets read, organized, and kept current, instead of sitting in a folder waiting for somebody's Saturday. Meeting preparation assembles itself from records the practice already holds. Drafting happens in the advisor's voice and waits for the advisor's judgment. Reconciliation just happens.
Run the math. If even half the non-client hours come back — and over time I expect more than half — the advisor gets those hours back to spend as they choose. Some will put all of it into more families: the same relationship capacity, barely used today, carries two or three times the households without thinning any of them. Some will hold their client count and go deeper — the tax work they always referred out, the estate conversation that never fit, the second annual meeting they could never justify. Some will take Friday back.
That last one matters as much as the first. Nobody should argue that forty hours a week of client meetings is the ideal life; plenty of excellent advisors want a more varied week than that, and should have one. What changes is who decides. Today the calendar decides — the week fills with work that has to happen, and whatever is left over is what the advisor gets to choose. Putting that work on rails hands the decision back. Agency over the week is the return, and each advisor gets to spend it their own way.
The floor moves too
Raising the ceiling is only half of it, and the less interesting half. Three floors move at the same time, and each one matters more to a practice owner than the ceiling does.
The first is who can be served. For decades, tailored, high-quality wealth service has been rationed — not by intent, by cost structure. When serving a household properly takes that many hours of human labor, the math forces a threshold: below some level of assets, a household cannot be served well and profitably at the same time. So the industry serves them thinly, with a call center, a model, an annual check-in — or not at all. Change the cost of the packaging and the threshold moves. The household that was uneconomic to serve properly last decade can be served this decade, by an advisor who actually knows them, because knowing them is no longer buried under thirty hours of process. For a practice owner, this is not a charitable observation. It is the addressable market. Every family you had to decline because the math didn't work was a relationship you were qualified to have and couldn't afford to.
The second is who can deliver. A planner three years in — good instincts, no pattern library yet — working alongside a system that carries the firm's methodology, surfaces what they would have missed, and shows its reasoning, gives better advice in year three than they otherwise would have given in year eight. The specialist knowledge that used to live in a twenty-person firm's org chart becomes available to the person actually in the room. More planners and advisors clear the bar for tailored, actionable advice, and they clear it years sooner. Against a hundred-thousand-advisor shortfall, that is not a marginal effect. It is most of the answer.
The third is the one that decides whether a practice grows: who you can build a team out of. Growth has always been gated less by finding clients than by building people — hiring, training, and supervising well enough that work leaves the owner's desk without the standard leaving with it. Plenty of practices cap not at a hundred clients but at the first hire the owner couldn't afford to get wrong. When the methodology is written down and runs on rails, a new hire starts inside a system instead of inside the owner's head. Training compresses. Supervision stops depending on the owner's memory and becomes review of work that shows what it did. And the downside of a hiring mistake shrinks, because the system holds the standard. For an owner, that is the entire calculation: return on their own time, and how much risk is attached to every attempt to buy more of it.
The craft advantage of the big firm was never mostly craft. It was mostly staffing, and software collapses that difference from below.
When a constraint that shaped an industry's entire structure for forty years turns out to be software, the industry doesn't just get more efficient. It gets restructured. Practice size, who gets served, what a household pays for, where the talent goes — all of it was downstream of that one constraint.
What I don't know
I don't know how fast this happens. I don't know which firms adapt. I don't know where the models plateau. What I'll stand behind is direction: the work that has been rationing this profession is exactly the work agents do well, and that has never been true of any technology before this one.
The hundred-client ceiling was never a measure of how much an advisor has to give. It was a measure of how much the rails underneath them could carry. The rails just changed — and everything above them gets to change too: the ceiling for the advisor, the floor for every family that could never quite afford to be somebody's client, and the standard a small practice can hold without a twenty-person firm behind it.
Further reading: Kitces — why the second 100 clients are less profitable than the first · Kitces — 50 great clients · Kitces via FA-Mag — how advisors actually spend their time · McKinsey via WealthManagement.com — a 100,000-advisor shortage by 2034