TL;DR: A bookkeeper records what happened. A controller makes sure it was recorded correctly and closes the books on a reliable schedule. A CFO decides what to do about it. These are three distinct jobs with distinct skills, and the most common and most expensive mistake is hiring a CFO to fix a problem that is actually a controller problem, or asking a bookkeeper to carry work that requires a controller.
Bookkeeper, controller, and CFO are not three sizes of the same job. They are three different jobs that happen to share a subject. Once you see that clearly, most of the hiring confusion in growing organizations resolves itself.
The clean version is this. Your bookkeeper answers what happened. Your controller answers is it right. Your CFO answers what do we do now. Everything else is detail.
The bookkeeper: what happened
A bookkeeper records transactions. Contributions and receipts in, bills and payroll out, bank and credit card reconciliations, and the daily maintenance that keeps the general ledger a truthful record of events. The work is detailed, deadline-driven, and unglamorous, and an organization with a bad one is in constant low-grade pain.
The skill is accuracy and consistency. A strong bookkeeper codes transactions the same way every time, notices when something does not look like it usually looks, and asks rather than guesses. That last habit is worth more than it sounds.
What a bookkeeper is not built to do is design. Ask a bookkeeper to restructure your chart of accounts, to build an allocation methodology, or to decide how a complex transaction should be treated, and you have moved past the role. Many will attempt it out of loyalty, which is how organizations end up with a general ledger that is internally consistent and structurally wrong.
The controller: is it right
A controller owns accuracy, controls, and the close. This is the role most growing organizations are actually missing when they think they need a CFO.
The controller makes sure the books are not just recorded but correct: that revenue is recognized in the right period, that restricted funds are tracked properly, that accruals are made, that the balance sheet reconciles line by line rather than in aggregate, that the monthly close lands on a predictable day instead of whenever it happens to finish. They own the internal control structure and the separation of duties. They are usually the person who makes an audit go smoothly or badly.
Here is the diagnostic that matters. If your financial reports arrive late, or arrive on time but cannot be fully trusted, or change materially after the fact, you have a controller problem. Hiring a CFO will not solve it. The CFO will simply discover the problem, tell you what you already suspected, and then spend a year doing controller work at CFO rates while the strategic questions you hired them for sit untouched.
The CFO: what do we do now
A CFO owns the forward-looking and fiduciary work. Cash flow forecasting and runway. The budget and its reforecasts. Board and finance committee reporting. Banking and lender relationships. Capital structure and financing strategy. Financial modeling for major decisions. Risk. For nonprofits and churches, restricted fund strategy, audit readiness, and reporting that satisfies both a board and the donors behind it.
The distinguishing skill is judgment under uncertainty. A CFO is valuable precisely in the situations where the numbers do not produce an obvious answer: whether to take on debt, whether a program can sustain itself, what a growth plan actually implies for staffing, what happens if the largest funder does not renew. That work requires someone who has seen enough of these decisions to know which assumptions are load-bearing.
It is worth saying plainly that a CFO with nothing reliable underneath them is a CFO you are wasting. Forecasting rests on accurate history. Modeling rests on a chart of accounts that maps to how the organization actually operates. Board reporting rests on a close that finishes in time to matter.
The sequence that works
Build the base before you buy the judgment. In most cases the right order is bookkeeper, then controller, then CFO, with the caveat that any of the three can be fractional or outsourced rather than employed.
A practical way to think about the trigger points, recognizing that they vary by complexity and not just by size:
- Bookkeeping is needed from day one and never stops being needed.
- Controller capability becomes necessary when transaction volume, restricted funding, or entity complexity grows past what careful bookkeeping can hold on its own, and when leadership starts depending on the monthly numbers to make real decisions.
- CFO capability becomes necessary when the decisions in front of leadership carry more weight than the current numbers can carry: a major capital decision, a funding concentration risk, a transition, an acquisition, or a growth plan that has to be tested rather than hoped through.
The exception worth naming is the transition. When a founder retires, a CFO departs, or a merger closes, the sequence compresses. You may need senior judgment immediately, before the base is solid, because the decisions will not wait. That is a real situation and it calls for a bridge, not a permanent structure built in a hurry.
Fractional versions of all three
Each of these roles can be bought fractionally, and the logic differs by role.
Fractional bookkeeping is common and works well, because the work is continuous but often part-time at smaller scale. Fractional controller capability is underused and frequently the highest-value option, because the work is genuinely episodic: heavy at close, lighter mid-month, heavy again at audit. Fractional CFO work fits the shape of the role almost perfectly, because executive financial judgment concentrates around budget season, board cycles, and specific decisions rather than spreading evenly across every week.
The advantage of buying more than one of these from the same firm is that the handoffs stop being your problem. The most common failure in a stitched-together finance function is not incompetence in any single role. It is the seam between them, where the bookkeeper assumes the controller caught it and the controller assumes the CFO was told.
The Counter-Move
When the numbers are not working, the instinct is to hire up. Bring in someone more senior, more credentialed, more expensive, and trust that seniority will resolve the confusion.
The counter-move is to diagnose the altitude of the problem before you buy the altitude of the solution. Late reports are not a CFO problem. Reports you do not trust are not a CFO problem. A chart of accounts that cannot answer your questions is not a CFO problem. Those are all structural, and structure is a controller-level fix at a controller-level cost.
Buy the CFO when the question is genuinely a judgment question. Buy the controller when the question is whether you can trust the answer. Most organizations under $25M that think they need the first actually need the second, and discovering that after the hire is an expensive way to learn it.
An invitation
If you are not sure which of these three your organization is missing, that uncertainty is itself useful information. It usually means the finance function grew by addition rather than by design, which is how nearly every growing organization gets here.
Novum builds the full finance ladder under one firm, from bookkeeping through fractional CFO, so the roles are sized to the organization rather than assembled from whoever was available. If you want a straight read on which level you actually need, we would be glad to have that conversation.