Key takeaways
- A controller owns accuracy. A CFO owns the decision. They are different jobs, and one is a prerequisite for the other.
- A CFO working from an unreliable close produces confident wrong answers, worse than no answer.
- Most businesses between $5M and $25M need a full-time controller and a part-time CFO. Many buy the reverse.
- The honest test is whether your problems are "the numbers are late or wrong" or "the numbers are fine and I still cannot decide."
A useful shortcut before any of the detail: a controller owns accuracy; a CFO owns the decision. The controller's question is whether the numbers are right. The CFO's question is what the numbers mean and what to do next.
That distinction sounds academic until you watch it go wrong. A business hires an expensive CFO because growth feels out of control, and six months later that CFO is personally rebuilding the reconciliations, because you cannot forecast from a close you do not trust. The business has bought a strategist and deployed them as a bookkeeper, at four times the rate.
The two roles, concretely
| Controller | CFO | |
|---|---|---|
| Owns | Accuracy | The decision |
| Core question | Are the numbers right? | What do they mean, and what do we do? |
| Typical work | Month-end close, reconciliations, reporting, controls, compliance calendar, AP and AR oversight | Forecasting, pricing, capital structure, capital allocation, lender and investor relationships, transactions |
| Time horizon | Last month | Next eighteen months |
| Fails when | The close slips or the numbers cannot be relied on | Decisions get made on instinct or too late |
| Needed from | ~$2M revenue, earlier if complex | ~$5M revenue, earlier if capital-intensive |
| Full-time at | Usually the first finance hire | ~$25M–$40M revenue |
Note the ordering in that final row. The controller is almost always the first finance hire, and the CFO is almost always the second. The sequence matters because the CFO role depends on the controller role having been done.
The diagnostic: which problem do you have?
Read the two lists below and count which side you recognize more of.
You have a controller problem if:
- The month-end close takes more than three weeks, or slips regularly
- You do not fully trust the numbers when you see them
- Your last audit produced a long list of adjustments
- Nobody can tell you your gross margin by product, contract or location
- Reconciliations exist but nobody independent reviews them
- Filing deadlines are met by heroics rather than by calendar
You have a CFO problem if:
- The numbers are accurate and timely, and you still cannot decide
- You are profitable but cash is unpredictable and nobody can explain why
- You find out about a covenant breach after the quarter closes
- Pricing decisions are made by feel because true margin is unknown
- A transaction is coming and nobody internal has done one before
- You have a budget, but no rolling forecast and no scenario view
If both lists feel familiar, you have a controller problem. Fix the foundation first. A forecast built on unreliable actuals is not a forecast, it is a guess with a spreadsheet attached.
The events that actually trigger a CFO
Revenue bands are a rough guide. In practice, the decision is usually forced by one of six events:
- A lender relationship with covenants. Covenants need forecasting forward, not reporting backward. Discovering a breach after quarter end removes every option you would have had eight weeks earlier.
- A planned transaction. Sale, acquisition, buyout or raise. Preparation started twelve to eighteen months out is worth real money in the outcome; preparation started at the letter of intent is damage limitation.
- A business model change. A second location, a new product line, a move from project to recurring revenue. Each changes the cash cycle in ways that are obvious afterward and invisible in advance.
- Cash that stopped tracking profit. The classic signal that growth is being funded out of working capital nobody is modeling.
- Outside capital arriving. Investors expect a reporting standard and a forecasting discipline that most owner-operated businesses do not have on day one.
- A decision you keep deferring. There is usually one: a price change, a hire, a piece of equipment. Deferring it because you cannot model it is itself the answer.
Why fractional works for this particular role
CFO work is genuinely lumpy. Budget season, a financing process, a transaction or a pricing review are intense; the months between are a results review, a forecast update and a management meeting. That shape suits a fractional arrangement in a way that controller work, which is relentless and monthly by definition, does not.
For a business between $5 million and $30 million, one to three days a month is usually the right cadence. That covers a monthly results and forecast review, a management or board meeting, and standing availability for decisions in between, with intensity rising temporarily around the set-pieces.
The question to ask a prospective fractional CFO
"What will you do in month one, and what will I have at the end of it?" A good answer is specific: a review of the last twelve months, a first cut of the forecasting model, a 13-week cash view, and a written assessment of where decisions are currently being made without information. A vague answer about "partnering strategically" is a signal to keep looking.
Can one person be both?
In a smaller business, yes, but be clear-eyed about which role wins when the week gets short.
The close has a deadline and the forecast does not. That asymmetry means controller work will always displace CFO work, every time, without anyone deciding it should. If one person holds both roles, the strategic time has to be protected explicitly (blocked in the calendar, with a deliverable attached), or the business ends up with a well-run close and no forward view at all, which is a perfectly comfortable place to be right up until it is not.
The sequence that usually works
- Stabilise the close. Reliable, on time, reconciled. Whether that is an internal controller or an outsourced accounting function matters less than that it is genuinely reliable.
- Add fractional CFO capacity. Forecast, cash view, KPI reporting, and the specific decisions that are currently being deferred.
- Scale the internal team. As volume grows, bring transactional work in-house where it is cheaper to run internally.
- Hire a full-time CFO when there is genuinely enough strategic work to occupy one, and use the fractional CFO to help you specify and assess the hire.
A fractional engagement that ends because the business outgrew it is a success, not a loss. Part of the job is building the internal capability to eventually not need us.
If you are unsure which of the two problems you have, that is a thirty-minute conversation rather than a research project. Bring your last three months of reporting and the decision you have been putting off. Those two things usually settle it. Our outsourced CFO service sets out what an engagement actually involves.