Do I need a CFO or a controller?

A controller makes sure the numbers are right. A CFO decides what to do about them. Those are different jobs, not two rungs of the same ladder, and most owners who feel their finance function is failing them have hired one and are waiting for the other. If your books are accurate and you still cannot answer what should we do next quarter, you do not have a controller problem.

The confusion is understandable. Both roles live in finance, both are senior, and on an org chart one appears to sit above the other.

What a controller does

A controller owns the record. Closing the month, reconciling the accounts, running payroll properly, making sure the statements are accurate and arrive on time, keeping the filings clean.

The output is reliability. When a controller is doing their job well, you stop thinking about whether the numbers are right, because they are.

That is not a small thing. A business without a functioning controller cannot make a good decision even with the best CFO in the country sitting beside it, because every decision is built on a number nobody trusts. This is why the order matters: the record comes first.

What a CFO does

A CFO owns what the numbers mean and what happens next. Which of your product lines is actually carrying the business. Whether the growth you are seeing is worth what it costs to fund. When to raise, when not to, and what a lender will ask before you go. What to stop doing.

The output is judgement, and it is directional rather than administrative. A CFO looks forward at things that have not happened yet; a controller looks back at things that have, and makes sure the account of them is true.

Neither is more important. They fail differently. A weak controller produces confident decisions built on wrong numbers. A missing CFO produces accurate numbers nobody acts on.

The test that separates them

If you are unsure which one your business is short of, ask yourself which sentence is closer to true:

"I do not trust the numbers." → That is a controller gap.

"I trust the numbers and I still do not know what to do." → That is a CFO gap.

Almost every owner can answer that in a second, and the answer is usually the opposite of what they were about to hire for. Businesses under pressure tend to reach for more accounting, because accounting is legible and quantifiable and you can put it in a job description. Direction is harder to specify, so it gets postponed.

A second, blunter test: when something in the numbers looks wrong, who is expected to raise it before you ask? If the answer is nobody — if bad news travels only when you go looking for it — that is a seat that is empty regardless of how many people work in finance.

Why "our accountant handles it" usually is not the answer

Your accountant or bookkeeper is generally doing controller-adjacent work, often well, and frequently as an external provider on a compliance calendar.

But an external accountant is engaged to produce a correct record and meet deadlines. They are not engaged to sit in your management meeting and argue that a customer segment should be dropped. That is not a criticism of accountants — it is a description of what they were hired to do. Asking them to supply strategic direction is asking for something outside the scope you agreed and are paying for.

This is also why "promote the bookkeeper" so rarely works. The skill that makes someone excellent at keeping a record accurate is precision under known rules. The skill that makes someone a good CFO is judgement under uncertainty. Some people have both. Most have one, and being asked to supply the other is a bad week for everybody.

We wrote about the moment this usually breaks: when founder-run finance breaks at scale.

When a fractional CFO makes sense, and when it does not

It makes sense when the decisions are real but not constant. You have accurate numbers and no one senior interpreting them. You are heading into something specific — a raise, a lender conversation, a pricing change, an acquisition, a year where growth has stopped turning into profit. That last one is common enough that we wrote it up separately: revenue growing but margin isn't.

It does not make sense when the record is not reliable yet. A part-time senior person cannot make good calls on numbers that are wrong, and will spend the engagement fixing bookkeeping at a senior rate. Fix the record first. Anyone who tells you otherwise is selling you the more expensive seat because it is the more expensive seat.

It also does not make sense if nobody will act on the output. A seat that produces recommendations into a vacuum is a cost with no return, and the honest thing is to say so before starting rather than after.

What the seat should look like on paper

Whatever you call it, insist on being able to draw its edges:

• Defined days. A specific number per month, not "as needed", which becomes either nothing or everything.
• Written scope. What this person decides, what they recommend, and what stays with you.
• What it explicitly is not. Not the audit. Not tax advice. Not bookkeeping. A seat you cannot draw the edges of is not a seat.
• No long lock-in. If the arrangement only works because leaving is difficult, it was not working.

The fractional CFO page covers what the seat costs and how the days are structured. If you are weighing several seats at once, the fractional overview sets out how they differ.

The short version

Ask what is actually missing. If it is accuracy, hire a controller and do not overpay for strategy you cannot yet rely on. If it is direction, adding more accounting will not produce it, however good the accounting is.

Most businesses need both eventually. Very few need both on day one, and the order is almost always the record first.

How to tell which one you actually need

Listen to your own questions for a fortnight. Write them down as they come. If most of them are about what already happened — did that invoice go out, why is this account off, where did the month land — you need a controller, and hiring a CFO will not fix it. If most are about what to do next, the reverse. Nobody needs to be paid to tell you which pile is bigger.

What needs someone who has done it before: building the first forecast that survives contact with reality. The second one you can maintain yourself; the first one is where the assumptions get argued out, and that argument is the value.

The decision that stays with you: whether you want the answer monthly or once. That is a question about how you like to run the business, and it changes the shape of what you should buy.