Where new roles actually come from
In a family office, new roles are rarely designed. They accrete. The assistant who booked the principal's travel starts handling the family's calendars, then the household invoices, then the insurance renewals. The CFO picks up IT because someone had to. The estate manager ends up running security because the guard company needed a point of contact and he was the one standing there.
None of this is a failure. It is how small, trusted teams work, and for years it is genuinely efficient. The problem arrives later, when the pile gets heavy enough that someone says the sentence every search begins with: we need to hire a person for all of this.
Why hiring against a pile fails
At that moment, most offices write a job description by listing what the last person did, or what three people currently do between them. What goes to market is not a role. It is a person-shaped hole.
The search that follows fails in predictable ways. Candidates match the title but not the pile. The strongest people read the description and see three jobs stapled together, and decline politely, because senior operators know what accreted scope does to a first year. The people who do accept often discover the full pile only after they arrive. And because the seat was never really defined, the office prices it against the title rather than the work, which is how a role carrying chief of staff duties gets offered house manager money.
The wreckage shows up on the other side, across searches that fail. The pattern is rarely a bad candidate. It is a seat that was never built.
Separate the seat from the person
The fix starts with one discipline: describe what the seat does, not what the people did. Take the full pile of tasks that triggered the hire and sort every item into three lists.
- Belongs to this seat. Work that is genuinely one mandate, done by one person, at one level of seniority.
- Belongs to another seat. Work that only landed in the pile because of who happened to be nearby. Return it to where it belongs, even if that seat does not exist yet either.
- Should stop happening. Every pile contains work that survives only because nobody ever asked why. A new hire is the one moment you can retire it without a fight.
Expect the third list to surprise you. It is rarely empty.
The one-page role charter
What remains in the first list becomes the role. Before any search starts, write it down on one page. Not a job description for the market. A charter for the family. Five sections, none longer than a few lines.
The mandate
Why the seat exists, in one sentence. If the sentence needs the word and more than once, you are probably looking at two seats. A real mandate reads like: this seat exists so that the family's properties run without the principal's involvement.
Decision rights
Three short lists: what this seat decides alone, what it recommends for someone else to decide, and what it must never decide. This is the section that prevents the quiet failure mode of family office roles, a senior hire with responsibility for everything and authority over nothing. It is a documented exit driver: on the chief of staff seat, rouka's retention data lists a scope-versus-authority mismatch first among the reasons hires leave.
The reporting line
One line, and a true one. Not the org chart as drawn, the org chart as lived. If the seat will really answer to the principal on Tuesdays and the COO the rest of the week, write that down now, because the candidate will discover it in month two either way.
First-year outcomes
Three things that will be visibly true after twelve months if the hire worked. Outcomes, not activities. Every vendor contract renegotiated or retired is an outcome. Manages vendors is a task.
The fence
What this seat does not do. This is the section that stops the accretion cycle from starting again. It is also the kindest thing you can hand a new hire, because it gives them permission to say no with your authority instead of their own.
The two-jobs test
Some piles refuse to become one role because they are not one role. The signs are consistent:
- The work serves two different masters with different definitions of success, the principal's personal life on one side and the office's operations on the other.
- The skills come from two different talent markets that barely overlap. People who are excellent at capital projects are rarely the people who are excellent at investment reporting, and the ones who claim both are usually strong at one.
- The compensation bands are far apart. When half the role prices at one level and half at another, one search cannot buy both. You will either overpay for the junior half or lose the senior half in a year.
When two or more of these are true, split the role. Two well-scoped searches are cheaper than one impossible one, and dramatically cheaper than the failed hire the impossible search produces.
Price the seat you designed
Compensation follows scope, not titles. Once the charter exists, price the seat against the market for what it actually does. The Benchmark carries the bands we price live searches with, by role and market, and the compensation guide covers how family office packages are structured above base.
One warning from the data: scope creep is compensation drift. A role that quietly absorbs a second mandate without a comp review is a resignation being scheduled. The charter is not just a hiring document. Reread it once a year against what the seat is actually doing.
Where a search partner comes in
Most of the work above does not require us. It requires an afternoon, one page, and the willingness to disappoint whoever wanted the pile to be one convenient hire. Where a partner earns its fee is in pressure-testing the charter against the market: whether the seat as designed exists in the wild, what it truly pays, and who is doing it well inside comparable families. If you are designing a seat right now and want that pressure test, that conversation costs nothing.
©2026 Talent Gurus Inc. All rights reserved.