A company that appoints a shared leader has bought two or three days a week. Over ninety days that is about twenty-six working days. A full-time appointment gets sixty-five in the same period and a six-month grace period on top of it.
So the engagement has to be run differently, and most of the difference is the company's job rather than the leader's.
Heidrick's 2026 survey of independent executives asked what clients get from them. The top three answers were filling a critical skill gap at 75%, bringing objective insight at 63% and accelerating a key initiative at 61%. All three depend on access. None of them survive a leader who spends the first month waiting for a system login.
Day one. The announcement
The highest-return act in the whole engagement takes ten minutes.
Announce the appointment internally, in writing, in the same terms a full-time appointment would be announced. Name what the leader owns. Name who reports to them for it. Name the days they are in the building. Say that decisions in that area now go through them.
A leader introduced as "someone helping us out with finance for a while" will be treated that way for the whole engagement, and no amount of competence recovers it. This is the single most common way a good appointment turns into an expensive advisory relationship.
Days one to five. Access, not briefing
Give the leader the things that take a week to get and cost nothing to grant.
- The accounting system, the CRM, the board pack for the last four meetings, the last audited accounts, the org chart with reporting lines and salaries where the seat needs them.
- A calendar with the recurring meetings they should be in, already accepted on their behalf.
- An email address on the company domain. A leader writing to the company's bank from an outside address spends the first month explaining who they are.
The instinct is to spend week one briefing the leader on the company's view of the problem. Resist it. The objective insight in the Heidrick number is the thing being bought, and it does not survive being told the answer first.
Weeks one to three. The diagnosis
The leader's job here is to meet people and read the record. The company's job is to open doors.
Four groups matter: the chief executive, everyone inside the function the leader is taking over, the two functions that depend on it most, and the external parties that touch the metric. The auditor, the bank, the lead agency, the principal regulator, whichever apply.
At the end of week three the leader should give the chief executive a written diagnosis: what the metric is actually doing, what is causing it, and the three things they intend to change first. One or two pages, not a deck.
Weeks four to eight. The first visible decision
Something has to change where the team can see it.
Not a policy document. A decision. Kill the reporting pack nobody reads. Move the month-end close. Fire the agency. Rewrite the pricing. Whatever the diagnosis pointed at.
The reason is not momentum for its own sake. It is that a fractional leader's authority is granted in practice rather than in the contract, and it is granted the first time they take a decision and it holds. A leader who spends eight weeks producing analysis has spent their entire credibility window proving they are a consultant.
By week six, the company should be able to say what has changed. If it cannot, the conversation to have is about scope and access, in that order, and it should happen at week six rather than at week twelve.
Weeks nine to twelve. The exit, designed early
This is the part almost nobody does, and it is the part that decides whether the engagement compounds.
A shared leader is in the building part of the week and will one day not be there at all. So from week nine, every process the leader runs should be being handed to a named person or written into a documented rhythm. The forecast has an owner who is not the fractional CFO. The pipeline review runs whether or not the fractional CRO is in the room.
An engagement that ends with the knowledge walking out was scoped without an exit. An engagement that ends with a promoted internal successor and a written operating rhythm has paid for itself twice.
Day ninety. The review
One question: what happened to the metric named in the contract.
Three honest outcomes.
It moved. Extend, and consider widening the scope or adding a day rather than starting a second search.
It did not move and the diagnosis explains why. Usually authority, data or a dependency outside the seat. Fix that and extend. Heidrick's data shows 42% of independent engagements run past six months and 16% past a year, so a first extension is the normal case rather than a concession.
It did not move and the diagnosis does not explain why. End it, at one month's notice, which is why the notice period is in the contract.
The thing that makes all of it work
Treat the leader as a member of the leadership team who is in the building two days a week, rather than as a supplier who is in the building two days a week.
Everything above is downstream of that single decision, and it is made on day one, in the announcement.