"I think we need a CFO."
It is one of the most common sentences I hear, and most of the time the person saying it is describing a different job entirely. Not because they're confused about their own business — they know exactly what's going wrong — but because the finance titles all sound like each other, and in the SME market they're used loosely enough to be nearly meaningless.
An "accountant" might be a bookkeeper, an outsourced firm, or a chartered accountant. A "finance manager" at a fifteen-person company is often doing controller work; at a two-hundred-person company the same title might mean payroll and invoices. Meanwhile "CFO" gets applied to anyone senior enough to be trusted with the bank token.
So ignore the titles. Compare the jobs.
The Three That Get Confused
A bookkeeper produces a correct record of what happened. Transactions coded, reconciliations done, GST filed, statutory accounts prepared on time. This is not junior work — a good one saves you from a great deal of expensive tidying later — but it is fundamentally backward-looking. It answers what happened?
A financial controller makes the record trustworthy and timely. They own the close, the controls, and the reporting calendar. When the numbers arrive within two weeks and nobody argues about whether they're right, a controller is why. They answer can we rely on this?
A CFO changes what happens next. Pricing, capital, capacity, which business to be in and which to leave. They answer what should we do? — and, more usefully, what will it cost us if we don't decide?
Those three questions are not seniority levels of the same job. They are different jobs, and a person who is excellent at one is frequently mediocre at another. The best controller I've worked with had no interest whatsoever in strategy. That was not a shortcoming; it was why the close never slipped.
It Isn't Skill. It's Temperament.
Here is the part the job descriptions never say out loud: these roles attract different kinds of people, and the traits that make someone excellent at one are frequently the traits that make them miserable in another.
Watch what each does when a number is missing.
The bookkeeper waits for the invoice. That is correct behaviour — the record should reflect reality, and an invented figure is worse than a gap. The controller chases the invoice, because the close is on the tenth and a gap is somebody's fault. The CFO writes down an assumption, labels it clearly as an assumption, and carries on — because the decision is due on Thursday and waiting has a cost too.
Same missing invoice. Three instincts. None of them wrong.
If you like the shorthand, DISC puts the first two firmly in high-C territory: conscientious, precise, allergic to sloppiness — with the controller carrying enough D to enforce a deadline on people more senior than they are. That last part matters more than most job ads realise. A controller who cannot bear to be unpopular on the fifth of the month is not a controller, whatever the title says. CFO work leans the other way, toward D and I, because a good half of it is persuasion — getting a founder to accept a number they'd rather not, or a lender to accept one they'd rather not.
The MBTI crowd says the same thing with different letters: the dutiful, detail-holding types gravitate to the ledger, the pattern-and-possibility types to the plan. I wouldn't hire on four letters — no code has ever told me whether someone can run a close — but the observation underneath is sound. The ledger rewards convergent thinking, where there is one right answer and the work is finding it. The forecast rewards divergent thinking, where there are five plausible futures and the work is choosing one and owning the choice.
Left brain and right brain is the version you'll hear at networking evenings. It's a caricature — no job lives in one hemisphere — but it endures because it points at something real: one seat is rewarded for closing questions, the other for opening them.
There's a tell in what each one finds satisfying. A bookkeeper gets a small, clean hit of completion from an account that reconciles: a finished thing, in a job where most things are never finished. A controller enjoys a system that holds under pressure — the month the auditors came early and nothing moved. A CFO enjoys being right about something that hadn't happened yet, which is a considerably less comfortable way to live.
So the useful hiring question isn't what someone can do. It's what part of the month they look forward to. If a candidate lights up describing a clean reconciliation, you have found a superb controller, and you should stop interviewing them for a CFO seat. If they light up describing a pricing argument they won, the reverse. Hire the temperament and the skills tend to follow. Hire the CV and you spend a year wondering why a talented person seems so unhappy.
The Comparison, in One Table
The table adds two options that aren't really on that spectrum at all. A full-time CFO is the same judgement made permanent — right when the big decisions arrive weekly rather than twice a year. And a scoped project isn't a seat: it's a defined change with an end date, which is also the shape that grant co-funding is built around.
Two things in that table surprise people.
The first is that a fractional CFO can cost less than a controller. That isn't a discount on seniority; it's a consequence of buying a few days rather than twenty-two. You get judgement at the level you need it, and you don't get someone available at 4pm on a Wednesday to chase an invoice.
The second is that the cheapest row is the one most businesses under-invest in. Bookkeeping is where the whole structure rests. Skimp there and everything above it inherits the mess.
The Order Matters More Than the Choice
Records, then reliability, then decisions. That is the sequence, and it does not bend.
The most common expensive mistake in this market is buying a CFO — fractional or otherwise — on top of books nobody trusts. The engagement then spends its first months doing controller work at CFO rates, everyone feels the value isn't there, and the founder concludes the whole category is overrated. It isn't. The order was wrong. I wrote about that failure and four of its cousins last week.
The same rule applies to software, incidentally. A reporting tool sitting on unreliable inputs doesn't give you insight; it gives you the same uncertainty, faster and in colour.
Records, then reliability, then decisions. Buying out of order is not a shortcut — it's the long way round with a bigger invoice.
The Pairing Most SMEs Actually Need
Framing this as a single choice is where the reasoning usually goes wrong. For most Singapore SMEs between roughly S$3 million and S$30 million in revenue, the right answer is not one seat. It's two, at different levels.
Underneath: a reliable, relatively cheap reporting layer — your outsourced firm or an in-house finance manager, depending on volume, with whatever controller discipline the size of the business justifies. Its job is that the numbers are right and on time. That layer should be as boring and as inexpensive as it can be while still being trustworthy.
On top: a few days a month of senior judgement that reads those numbers and turns them into decisions. Pricing that holds. A forecast the bank believes. An honest answer on whether the second location is a good idea.
The mistake is buying one seat that tries to be both. A single hire in the middle of the range — senior enough to advise, cheap enough to also do the close — usually ends up doing the close, because the close has deadlines and strategy doesn't. Urgent work drives out important work every time. Two layers, priced differently, is how you stop that happening.
It's also why a fractional CFO shouldn't be pitched as a replacement for your accountant. If someone offers to be both, ask which one gets dropped in a busy month. There's only one honest answer.
The One-Sentence Test
Before you advertise a role or sign an engagement letter, finish this sentence honestly:
In twelve months, this person will have made ______ different.
If the blank fills with "our books will be accurate and our filings on time", you want a bookkeeper. If it fills with "the numbers arrive by the tenth and I stop arguing about them", you want a controller. If it fills with "we'll have chosen between three growth paths, with the arithmetic behind the choice", that's a CFO — and if that choice happens twice a year rather than twice a week, fractional is the honest shape.
If the blank refuses to fill at all, you have found something more useful than a shortlist: you've found out you're not ready to hire yet, which is a much cheaper thing to discover now than eleven months from now.
If you'd like to fill that blank out loud with someone who has sat in most of these seats, that's what the first conversation is for.