First finance hire vs fractional CFO: when to switch

Bookkeeping is the recording function: transactions coded, bank and card accounts reconciled, payables and receivables kept current, payroll journals posted. Accounting is the interpretive layer built on top of it — accruals, revenue recognition, period cut-off, and producing statements someone can rely on. Confusing the two is the most common reason an acquired business has clean-looking books that cannot answer a question.

A controller owns the close and everything that has to be true for the close to be trustworthy: the account structure, the reconciliation discipline, the definitions behind the numbers, the reporting package. This is the rung with the best value-for-money in most acquired businesses of this size, and it is chronically skipped because it is the least marketed of the four.

A fractional chief financial officer brings judgement that is episodic by nature — forecasting, capital structure, lender negotiation, pricing economics, the decision to take on more debt or not. Episodic judgement is exactly what fractional engagement suits. It is not a substitute for the controller function, and the failure mode of hiring one first is that a senior person spends their days fixing a chart of accounts.

A full-time chief financial officer becomes coherent when the judgement stops being episodic — multiple entities, real capital-markets activity, an acquisition programme — and not before. Most businesses in the ten-to-a-hundred headcount band never reach it, and that is a fine outcome rather than a failure to scale.

There is a diagnostic that resolves this question faster than any comparison of titles: write down the three questions you most need the finance function to answer next quarter, then ask which rung answers them. 'Are the books accurate and current' is a bookkeeping-and-cleanup answer. 'What is the margin on each service line and which customers are unprofitable' is a controller answer, and it is usually blocked behind an account structure that has to be redesigned first. 'Should we refinance, and what does the business look like at twenty per cent less revenue' is a chief-financial-officer answer.

In an acquired business the first two dominate for a reason. The inherited ledger is typically a single-user desktop file with commingled personal expenses and a chart of accounts that grew by accretion, and the cleanup plus redesign is a defined project with an end date rather than an ongoing role. Scoping it as a project — and hiring for it as one — is cheaper and faster than hiring a person and hoping the project falls out of them.

Mapping where each number comes from is part of the same diagnosis. Some figures come from the ledger, some from the field-service or inventory system, some from the customer database, and some from a person typing. The volume of manual entry is a direct measure of how much of the finance function is currently a person rather than a system, and it tells you whether you are hiring for judgement or for data entry.

A business carrying acquisition debt has an obligation an unleveraged business does not: a reporting package, on a contractual date, in a specified form. That obligation is controller work, and it is not optional or deferrable. It raises the floor on the finance function well above what the same-sized business without debt would need, which is why generic advice pegged to revenue or headcount consistently under-specifies this seat.

It also adds a recurring analytical requirement. Coverage of debt service and fixed-charge coverage need computing monthly, before the lender computes them, and a thirteen-week rolling cash forecast built direct — receipts minus disbursements — is what tells you whether a payroll and a loan payment collide. None of that is chief-financial-officer work. All of it is controller work with a standing calendar.

The corollary is a sequencing rule. Get the close reliable and the reporting obligation met first, then buy episodic judgement. Buying judgement before the data is trustworthy means paying a senior rate for someone to form opinions about numbers that are wrong, which is the most expensive way to discover your chart of accounts needs rebuilding.

Write the outcomes before writing the advertisement. A job scorecard states what the role must have produced by a given date — close completed by the eighth working day, reporting package delivered ahead of the covenant date, margin by service line available monthly, a definition dictionary maintained — plus the competencies that make those achievable and how each will be measured. Then interview against it in a consistent structure rather than conversationally, because a conversational interview for a finance role selects for people who are pleasant to talk to about finance.

The inherited bookkeeper is the awkward case and deserves a fair one. They frequently know more about the business's customers, vendors and quirks than anyone else including the seller, and they are frequently not the person who can own a controller-grade close. Those two facts are compatible. The move that respects both is to keep the institutional knowledge in the role it fits and hire the capability above it, rather than promoting on tenure and then managing the consequences.

That is a specific instance of the general problem in an acquired business: the best technician is often a poor manager, and promoting them removes your best practitioner and creates your weakest supervisor. In finance the equivalent is promoting the bookkeeper who has been there nineteen years into a role whose core requirement is telling the owner something they do not want to hear about the numbers.