The best finance function redesigns start with the CFO job spec because that document sets the ceiling for everything below it. Get the spec wrong and you will hire a leader whose instincts point the whole function in the wrong direction, no matter how good the systems, headcount plan or board reporting pack you build around them.
Most boards approach it the other way around. They redesign the org chart, pick a new ERP, agree a reporting cadence, then write a spec that describes the person who fits the plan. That sequence looks logical and usually fails, because the CFO you hire will rebuild the plan anyway. A capable finance leader always reshapes the function to match how they think. So the real design decision is not the org chart. It is the person, and the spec is where you decide who that person is.
Why does the CFO spec set the ceiling for the whole function?
A finance function rarely rises above the ambition written into its top job. If the spec asks for a safe pair of hands to keep the accounts clean, you will get controls and clean accounts, and very little forward-looking work. If it asks for a commercial partner who can price deals and challenge the CEO, you get a different function entirely, one that pulls talent, tools and board attention behind commercial questions.
The spec also signals what kind of person will say yes. Strong candidates read specs closely. A vague or box-ticking spec attracts people who are comfortable being managed rather than trusted, and it repels the operators who could actually transform the function. You are advertising your ambition whether you mean to or not.
What does a redesign-led spec look like in practice?
Start with the decisions the business needs the CFO to own in the next eighteen months, not the tasks. There is a real difference. "Owns the refinancing and the covenant relationship with lenders" is a decision. "Manages the finance team" is a task, and it tells a candidate nothing.
We usually push boards to name three or four outcomes that will define success. Examples we see often in mid-market and PE-backed businesses:
- Build a forecasting model the board actually trusts, replacing the current spreadsheet no one believes.
- Get the business ready for a fundraise or exit inside two years, including a clean data room and a defensible equity story.
- Rebuild an FP&A capability that gives operators real unit economics rather than lagging management accounts.
- Take a founder-run finance function through its first proper controls and audit cycle.
Each of those outcomes implies a different hire. The refinancing CFO and the FP&A-building CFO are not the same person, and pretending one spec covers both is how boards end up disappointed nine months in.
How does the spec shape the team you build underneath?
Once the CFO outcomes are clear, the layer below almost writes itself. A CFO hired to drive commercial decision-making needs a strong Head of FP&A early and can live with a lighter controller function for a while. A CFO hired to fix a messy close and get audit-ready needs a serious Financial Controller first, and FP&A can wait.
This is where sequencing saves money. Boards that hire the CFO against a clear spec, then let that CFO build the next two roles, avoid the expensive pattern of recruiting a team that the new leader quietly dismantles. We have placed Financial Controllers and Heads of FP&A who were hired six months before the CFO, and in more than one case the incoming CFO restructured both roles because they were built for a plan they did not share. That is wasted money and wasted goodwill.
If you are shaping those layers now, our live finance leadership roles give a sense of how specific these mandates have become.
What goes wrong when the spec is generic?
Generic specs create three predictable problems. First, they attract a wide, weak field, because they give strong candidates no reason to self-select in. Second, they make interviews unfocused, since panels end up assessing likeability rather than fit against real outcomes. Third, they store up conflict, because the CFO and the board discover their different assumptions only after the offer is signed.
The most common failure we see is a spec that lists both "transform the commercial model" and "personally own the monthly close" as if one person will happily do both at the same seniority and salary. Those pull in opposite directions. A good spec is honest about the trade-off and about which one matters more this year.
How should a board actually build the spec?
Treat it as a board conversation, not an HR task. Get the CEO, the chair and the relevant investor in a room and force agreement on the two or three outcomes that justify the hire. Write those down before anyone touches competencies or qualifications.
Then pressure-test it against the market. A spec that asks for every skill at the top of the range in one affordable person is not a spec, it is a wish. This is where an experienced recruiter earns their fee, by telling you which combinations exist and which do not. We do this work at Harper May every week, and the honest conversation about trade-offs usually happens before we advertise anything.
The reward is a redesign that holds. When the spec is right, the CFO you hire builds the function you actually needed, rather than the one you thought you wanted.
Common questions
Should we redesign the org chart before or after hiring the CFO?
After, in most cases. Agree the outcomes and the CFO spec first, hire against them, then let the CFO design the team. Redesigning the chart first usually means paying twice.
How detailed should a CFO job spec be?
Detailed on outcomes, lighter on tasks. Name the three or four decisions the CFO must own and what success looks like in eighteen months. Avoid long generic competency lists that attract a weak field.
Who should own the CFO spec?
The board, led by the CEO and chair, with investor input where relevant. It is a strategic decision about the direction of the finance function, not a document to delegate wholesale to HR.
