Do I Need a Controller If I Already Have a Bookkeeper?

If you make real money decisions off your numbers, you probably do. A bookkeeper records what happened. Your controller checks whether it is right and tells you what it means. Between the two sits trust, and that gap is what costs you when a number is wrong. One missing signal and you can wait. Three and the hole is already open.

What does a controller do that a bookkeeper doesn't?

A bookkeeper hands you a finished set of books. Then a controller asks the question no one else in the process is paid to ask: are these books actually true before you act on them? Want the full ladder from recorder to strategist? The bookkeeper vs controller vs CFO breakdown maps all three roles.

Here is the kind of thing that slips. Say a recurring software subscription gets coded to a job instead of overhead. Small, easy to miss. But it pulls cost out of the place that shows your real operating margin and parks it in cost of service. On one engagement that single habit made a service line look about six points more profitable than it was. For months, the owner had been pricing new work off that inflated margin. Every transaction was recorded correctly in the sense that nothing was missing. Nobody asked whether it was right.

That is the whole difference.

A controller sits in that gap. They close the month, reconcile the judgment calls, and weigh the things a recording system will never flag: accrual cutoff, revenue booked in the right period, whether the result is even reasonable. Sometimes two errors cancel and the margin looks fine while both halves are wrong. Rare, but it happens, and a recording-only setup will never catch it. Done well, the report that reaches you is one you can act on without a second guess. So that is what you are buying. Not more entries. Confidence in the entries you already have.

How do you know you have outgrown bookkeeping alone?

Usually you feel it before you can name it, a low hum that the report and the business in your head are telling two different stories. You cannot say why. That feeling is the signal.

The patterns I see most:

• You do not trust your own financials. You get a report and quietly discount it. That is not a software problem. It is a missing review layer.
• The month never really closes. Books stay open, numbers shift after the fact, last month is still moving in week three of this one.
• You make real decisions off the numbers. Hiring, pricing, taking on debt, a distribution. The cost of a wrong number just went up.
• Complexity crept in. Multiple revenue streams, more staff, a line of credit, jobs that span months. More moving parts, more places a recording-only system gets it wrong. This is also when books fall behind and need cleanup.
• Year-end turns up accounting messes, not just tax positions. Some of what a CPA surfaces at year end is genuine tax work: depreciation, owner add-backs, return positions. But if they are also finding miscoded accounts and wrong-period entries nobody caught, that is the accounting review that never happened in between.

One of these and you can probably wait. Three and you are already paying for the missing controller. Just not to a controller. The bill shows up as mispriced work and decisions made off a number that was never real.

Cheaper ways to cover the review layer

A full controller on day one is not always the move. What you need is the review function. Below is a rough map of how owners cover it, cheapest to most, framed as what I see in practice, not a rate card.

Most $1-5M owners land in the middle two rows. What separates review from controller comes down to one word: ownership. A light review checks the bookkeeper's work and hands it back; a controller owns whether the close is right and carries that on their own shoulders. Buy the judgment, not the 40 hours. Pay for the layer you are actually missing.

How do you score yourself in two minutes?

Answer these honestly, the way you would if no one were watching and no advisor were in the room. Each yes is a point.

1. Do you quietly discount your own financials when you read them?
2. Does last month keep moving after it should be closed?
3. Have you made a hire, a price, or a debt call off these numbers in the last quarter?
4. Did year-end turn up accounting messes, not just tax adjustments?
5. Has the business gotten more complex than the books were built for?

Zero or one, you can wait. Two, watch it. Three or more, the controller-shaped hole is already costing you. The monthly reporting that founders can actually act on is the thing that closes it.

The honest answer

When you ask whether you need a controller, you are really asking something quieter. Can I trust these numbers. The bookkeeper is not the problem and never was; they are doing exactly the job they were hired for, which is recording what happened, transaction by transaction. Checking is the job nobody is doing.

So work backward from the trust question. Would you bet a hire or a price change on this month's report without flinching? Then your bookkeeper is enough. Hesitate, even a little, and that hesitation is the controller-shaped hole in your finance function staring back at you. Hire the review, not a second bookkeeper. A cheaper recorder cannot fix a missing reviewer.