Fractional CFO vs controller vs bookkeeper: who actually does what

Three job titles, used as though they were interchangeable. They are not. Owners end up hiring the wrong one, or paying for one and quietly expecting another.

Here is the plain version.

The bookkeeper records what happened

Every transaction categorized, every account reconciled, the month closed. If someone asks what you spent on materials in April, the bookkeeper is the reason there is an answer.

Good bookkeeping is not data entry. The judgement is in the structure. Whether the chart of accounts reflects how you actually make money. Whether costs land against the right job, location, or provider. Whether the close happens early enough in the month to still be useful.

Get that structure wrong and everything built on top of it inherits the error.

The controller owns whether the numbers are right

A bookkeeper produces the record. A controller is accountable for it being accurate, complete, and on time.

That means reviewing the close, catching errors before they reach a statement, enforcing consistent process month over month, and building reporting other people can rely on without checking it first.

In most businesses under roughly $10 million in revenue, this is not a separate person. Either the bookkeeper takes on more, or the CFO does, or the owner quietly does it on Sunday night.

The CFO decides what to do about the numbers

Forecasting. Pricing. Margin strategy. Capital decisions. Lender relationships. Getting the business ready for whatever is coming next.

A CFO does not make your numbers more accurate. That is the controller's job. A CFO makes them mean something.

The distinction matters because it explains a common disappointment. Owners hire a CFO expecting the reporting to get cleaner, and the reporting does not get cleaner, because that was never the job they bought.

The order you actually need them

Bookkeeping first. Always. It is the foundation, and nothing above it works without it.

Then, usually, a CFO. Most owner-operated businesses hit the point where the decisions get expensive well before they hit the point where they need a dedicated person policing the close.

The controller function tends to come last, or gets absorbed by the other two, until the business is large enough that errors are expensive and nobody has time to catch them personally.

How to tell which one you are missing

Match the symptom to the role.

Your books close late, or they close on time and you do not fully trust them. That is a bookkeeping problem.

You trust the numbers, but you still cannot tell which jobs, locations, or providers make money. That is also usually bookkeeping, specifically the chart of accounts. The data is accurate and structured in a way that cannot answer the question.

You have accurate, well structured numbers and still cannot decide what to do with them. That is a CFO problem.

Statements arrive on time, and every few months someone finds an error nobody caught. That is a controller problem.

The mistake that costs the most

Hiring a CFO on top of books nobody trusts.

It happens constantly. The owner knows something is wrong financially, assumes the answer is more senior help, and brings in a CFO to fix it. The CFO then spends the first two months rebuilding the books, which is not what either side signed up for, and produces confident recommendations resting on inputs that were never solid.

Fix the foundation first. It is cheaper, it is faster, and it makes everything after it worth paying for.

All three exist in fractional form

Each of these roles can be structured around what the business needs.

A business doing $3 million in revenue can have properly structured books, a real monthly close, and a CFO in the room for the decisions that matter. The roles can work together without becoming one job.

The question is which role you are actually missing. For a look at how we run an engagement end to end, see how we work.