Bookkeeper, Controller, or CFO: How to Tell What You Need
For owners and leaders of growing companies
Whether you have one employee or 40, it is the same question: which parts of this business actually make money, and what is quietly eating the rest? Most owners cannot answer that from the reports they are handed, so growth turns into a series of educated guesses.
The instinct is to hire a CFO. Then you see what a full-time one costs and the idea dies right there, which is the wrong place for it to die. What a company this size usually needs is CFO-level judgment a few days a month, sitting on top of books somebody is genuinely managing. That costs a fraction of a salary, and it happens to be the same work that makes the business worth more if you ever decide to sell. This is the order to build it in.
Three jobs, not three sizes of the same job
The gap is rarely that the bottom layer is failing. It is that nothing above it exists.
Signs you have outgrown the reporting you have
The conversation tends to start one of a few ways. Some of these come from the owner's side of the table:
"I have a bookkeeper and a CPA and I still can't tell you which jobs made money last year."
"We closed the month three weeks ago. I got the report Tuesday."
"I know we're profitable. I don't know why the account is empty."
And some come from the person who runs finance inside the business:
"I can get you that number. It will take me four days and three spreadsheets."
"I've asked for a controller twice. Both times I got asked what it would cost."
"I spend the first two weeks of every month closing and the last two explaining why the close took two weeks."
None of these are accounting problems in the way they sound, and none of them mean anyone is doing a bad job. They are the same structural problem wearing different clothes: the finance function that was exactly right for the business four years ago is still the finance function, and it is not the same business.
Nothing announces that threshold when a company crosses it. There is no failure, no missed deadline, nothing visibly broken. The reporting just quietly stops being able to answer the questions that have started getting asked. Which is why most companies sit past it for a year or two before anyone puts a name to it.
If you run finance inside the business
You already know most of what follows, and the diagnostic sections below will not tell you anything you have not already concluded. The sections worth your time are the one on systems and the one on what sits underneath you.
The pattern we see most often is not a finance leader who has lost the thread. It is a capable one personally carrying the close, the reconciliations, and the reporting because the bench underneath was never built, and then being measured on strategic work there is no room left to do. When that is the situation, the constraint is capacity, not capability. It is also an easier case to make to an owner or a board with something in writing behind it. Use this for that.
How to tell if your books are being managed or just kept
If you run finance in the business, skip this one. You already know the answer. If you own the business and are not certain, it takes about 10 minutes. Open your books and look for three things: prepaid expenses, accrued liabilities, and amortization schedules.
If your books are meant to be on an accrual basis and the annual insurance premium still hits the P&L as one lump in January, or there is no accrual for work performed but not yet invoiced, then they are being kept, not managed. If they are on a cash or tax basis by design, that is a different question, and the one that matters is whether you are making accrual-level decisions off a cash-basis picture.
This is the cleanest line between bookkeeping and controllership. The owners we talk to know their businesses and their people well, and the transaction-level work is usually handled competently. What is missing is everything past the basic transaction: the entries that make a monthly P&L actually reflect the month it belongs to.
Until those entries exist, every decision made off your monthly numbers is being made on a distorted picture. Not a wrong one, necessarily. A distorted one, and there is no reliable way to know by how much.
If that test came back clean, your issue is further up the stack and the rest of this guide will get you there faster. If it did not, you have just found the gap, and everything below is about what to do with it.
What a bookkeeper, controller, and CFO each actually do
You likely know the broad strokes here. The reason to be precise about it is that the boundaries are where the problems live. The titles get used loosely, especially at companies this size, where one person routinely covers all three and only has time to do one of them well.
A bookkeeper records what happened. Transactions in, transactions out, bank feeds reconciled, invoices sent, bills paid. The job is accuracy at the transaction level, the foundation of what we handle under day-to-day accounting.
A controller makes sure what was recorded is right and closes the books on a schedule you can plan around. Accruals, prepaids, reconciliations, internal controls, financial statements that tie out. The job is a trustworthy record.
A CFO decides what to do about it. Forecasting, cash strategy, pricing, capital structure, lender and investor conversations, whether to buy the building or lease it. The job is judgment applied to the record.
Notice the dependency. A CFO's advice is only as good as the controller's numbers, and the controller's numbers are only as good as the bookkeeper's records. That ordering is where the expensive mistakes cluster.
Bookkeeper vs controller vs CFO: who owns what
Where the work sits varies somewhat by company, but the boundaries below hold across most businesses between $2 million and $50 million in revenue. The useful exercise is not reading it top to bottom. It is finding the rows nobody at your company currently owns.
| Work | Bookkeeper | Controller | CFO |
|---|---|---|---|
| Recording transactions, AP and AR | ● | ○ | – |
| Bank and credit card reconciliation | ● | ○ | – |
| Payroll processing and compliance | ● | ○ | – |
| Sales tax filing and compliance | ● | ○ | – |
| Accruals, prepaids, amortization schedules | – | ● | – |
| Month-end close and close calendar | ○ | ● | – |
| Balance sheet reconciliation and review | – | ● | – |
| Financial statements that tie out | ○ | ● | ○ |
| Internal controls, segregation of duties | – | ● | ○ |
| Inventory and cost accounting | – | ● | ○ |
| Job costing and project accounting | – | ● | ○ |
| Multi-entity consolidation | – | ● | ○ |
| Audit prep and CPA coordination | – | ● | ○ |
| Budgeting and forecasting | – | ○ | ● |
| Cash flow planning and working capital | – | ○ | ● |
| Product, customer, and location margin analysis | – | ○ | ● |
| Pricing strategy | – | – | ● |
| Banking, lender, and investor relationships | – | ○ | ● |
| Financial systems selection and implementation | – | ○ | ● |
| Due diligence, recapitalization, sale readiness | – | ○ | ● |
| Board and investor reporting | – | ○ | ● |
● Owns it ○ Supports or reviews – Not this role
When do you need a controller?
Which of these are true?
- Your monthly close takes longer than 10 business days, or nobody can tell you how long it takes.
- Your P&L and your bank balance tell two different stories and no one can reconcile them quickly.
- Your books are effectively on a cash basis, but you are making accrual-level decisions.
- Nobody reconciles the balance sheet. Everyone watches the P&L.
- One person records the transaction, approves the payment, and reconciles the account, because there is nobody else to split it between.
- A lender, bonding company, or insurer asked for something that took days to produce.
- Reports are accurate and arrive too late to act on.
- The role has grown well past what it was scoped for when it was filled.
- A second entity, location, or revenue stream got added and the reporting structure never caught up.
Three or more, and the gap is real. It is almost never a performance problem. It is a scope problem: the job outgrew the seat it was built for.
That distinction matters, because this is the point where companies do the wrong thing. A bookkeeper who has run a $3 million company's books beautifully for nine years is not failing when the company reaches $18 million. The work changed underneath them. Adding controller-level oversight above that person is usually the right answer and is a completely different move from replacing them.
When do you need a CFO?
Which of these are true?
- Nobody can say quickly which products, customers, or locations actually make money.
- The business is planning something it has not done before: a sale, a recapitalization, an acquisition, a new facility, an ESOP.
- Cash is tight in a year that shows a profit and the gap has not been explained.
- Pricing has not been revisited since input costs moved.
- A bank or investor negotiation is coming and nobody in the room does those regularly.
- There is a forecast, and it is not what decisions actually get made on.
- Capital decisions are getting made without a model underneath them.
- The books are clean, on time, and correct, and the next move still is not obvious.
That last one is the real dividing line. Clean books and no direction is a CFO-level question. Unclear books are a controller-level question, and no amount of strategic input fixes them.
If three or more are true, the business is usually carrying CFO-level questions with nobody assigned to them. That work does not require a full-time hire. It is what fractional CFO and management reporting is for.
Is it your team, or your accounting system?
This is the question that gets skipped, and it is the reason a lot of good hires fail.
You can bring in an excellent controller, pay properly for the seat, and watch them spend their entire first year building spreadsheet workarounds, because the constraint was never the person. It was the software. Entry-level accounting systems are genuinely good, and every one of them has a ceiling.
Signs your system has become the constraint
- Core processes (inventory, order management, procurement) live in Excel rather than in the accounting system.
- The same information gets entered in two places by two people.
- You run multiple entities and your consolidated statements are assembled by hand every month.
- Inventory accuracy started degrading after you passed a few hundred SKUs or opened a second location.
- There is no native support for bills of materials, work orders, or landed cost, so those live outside the financials.
- Close time is spent compiling data rather than reviewing it.
- Every increase in order volume requires another person to handle the manual steps.
- You need approval workflows and permission controls the software cannot enforce.
Several of these true at once is a systems conversation, not a hiring conversation, or more often, both in the right order.
The honest answer is usually sequencing. Get the record trustworthy in the system you have, understand precisely where that system runs out of road, and then make the change deliberately rather than in a panic during a year-end audit or a due diligence request. That evaluation is its own piece of work. We cover it under system design.
We will work inside whatever you already use. We will also tell you plainly when the software is the thing holding you back, and help you move when the timing is right. Those are two different services, and conflating them is how businesses end up with an expensive implementation that solves a problem they did not have.
What accounting support looks like in your industry
What’s missing above the bookkeeping layer looks different depending on what you do. Generic advice is where most of this conversation goes wrong.
Manufacturing and distribution
The failure point is almost always cost. Material, labor, and overhead stop flowing cleanly into the cost of a finished unit. Freight, duty, and brokerage get expensed on receipt instead of landing in the product's cost. You buy in kilograms, stock in pounds, and consume in pieces, and the unit conversions quietly corrupt the math. Work-in-process produces variances nobody can explain.
The symptom you feel is simpler than any of that: you cannot say with confidence which products are still profitable after your suppliers raised prices. The monthly margin report looks fine in total while specific lines quietly lose money.
We wrote a longer guide on that specific problem: why your gross margin shrinks when nothing looks broken, with the arithmetic and five checks you can run yourself.
Real estate, construction, and development
Job costing and work-in-process schedules are the whole game. Costs land in the wrong job or the wrong period, over- and under-billing goes untracked, and the WIP schedule your surety or lender wants is reconstructed from memory each time it is asked for.
Add entity structure to that (one LLC per project, capital stack across several) and monthly reporting becomes a manual assembly job. The cost of getting it wrong is not an accounting problem. It is a deal that slows down or a draw that gets questioned.
Professional services
Revenue recognition and unbilled work are where it breaks. Labor is the dominant cost and it is hard to see clearly. Utilization is tracked in the billing system and never reconciled against the financials. Partner distributions get made on a cash view of a business that is not actually a cash business.
Owners of these firms know their business cold. What they usually don't have is reporting built to answer the questions they're actually asking: utilization, realization, what a matter or an engagement really earned.
Nonprofits
Fund accounting is a genuinely different discipline, and a bookkeeper who has only worked in for-profit environments will not know what they do not know. Restricted versus unrestricted, release from restriction, grant compliance and reporting deadlines, functional expense allocation across program, management, and fundraising. All of it has to be right before the audit, and all of it is invisible in a standard chart of accounts.
Most nonprofit leaders we meet have been burned once already by someone who treated the organization like a small business with donations.
Home services: plumbing, HVAC, and electrical
The money lives in two places most owner-operated trades businesses cannot see clearly: the true margin on service versus install versus maintenance work, and the balance of prepaid maintenance agreements sitting on the books. Both get managed on instinct until something forces the question.
The thing that usually forces it is an offer. Private-equity roll-ups are active in the trades right now, and the number a buyer lands on depends almost entirely on what the books can prove. We covered that separately: what buyers examine before they price a plumbing or HVAC business.
Private-equity backed and portfolio companies
The bar moves the day the deal closes. Monthly reporting packages on a sponsor's calendar, diligence-ready detail on request, add-on acquisitions that need integrating into a single chart of accounts. The finance function that carried the business to the transaction is frequently not the one that survives the first year after it.
Why a CFO hire stalls without a controller underneath
The most expensive version of this is adding CFO-level leadership without building the layer underneath it.
It happens constantly and the logic behind it is sound. The pain that gets attention is strategic, so the business goes and finds the most senior person it can. What follows is predictable and is nobody's fault in particular. That person inherits a close process that does not hold up, and now holds two full-time jobs: building the foundation, and doing the strategic work that was the entire reason for bringing them in. One of the two loses. It is almost always the strategy.
Fix the record before you buy the interpretation. Judgment applied to numbers nobody has validated produces confident, well-reasoned, expensive conclusions built on a picture that is off, which is worse than no conclusion at all, because the business acts on it.
The reverse mistake is quieter and more common: staying with bookkeeping-only coverage long after a second layer would have earned its keep, because nothing is visibly broken. Nothing is visibly broken because nobody is looking at the balance sheet.
None of which is an argument against CFO-level help. It is an argument about what has to be underneath it. Buy the judgment at the scale you can actually use (a few days a month rather than a full-time seat) and buy it on top of a close that holds. Done in that order, the same money produces answers instead of a second full-time job for whoever you hired.
Full-time, fractional, or outsourced accounting: how to decide
Once you know what level of work the business needs, there is a second question: how you get it. Cost is the obvious variable and the least interesting one. These matter more.
Coverage. A single in-house hire is a single point of failure. When that person takes leave or resigns, the function stops and the institutional knowledge leaves with them.
Pattern recognition. Someone who has spent 11 years inside one company knows that company extremely well. Someone who has worked across dozens has seen your situation before (the inventory problem, the system migration, the diligence request) and is not solving it for the first time on your dollar. Our team has worked across enough accounting systems and enough situations that a problem you have never seen before is usually one someone here has.
The right seniority mix. Most companies in that range need controller-level work every week and CFO-level thinking a few days a month. Hiring that as full-time headcount means either paying for idle senior capacity or underhiring and hoping.
Flex in both directions. Complexity moves. A recapitalization, an acquisition, or an audit year needs more for a while and then does not. Headcount does not flex; a scoped relationship does.
When you don't need a controller or a CFO
Worth saying plainly, because most guides like this one are constructed to conclude that you need the author's services.
If you are early-stage with minimal revenue and no real complexity, you need a good bookkeeper. Not a controller, not a fractional CFO. A competent bookkeeper and a tax CPA will carry you a long way, and paying for senior financial expertise before there is anything complex to apply it to is money spent on capability you will not use.
If the first question is what it costs rather than what it would change, the timing is probably not right yet. Outsourced finance is bought for expertise and continuity, and if the value is not yet obvious, the engagement will feel expensive no matter what it costs, which is worth knowing before either side spends time on it.
If nobody on your side can be reachable for monthly decisions, no outside firm can fix that. More on this below, because it is the single most reliable predictor of whether one of these relationships works.
If you want a one-time cleanup and nothing after it, hire for the project. A cleanup is a legitimate piece of work and sometimes the right door in. But if the project is genuinely all you want, an ongoing relationship is the wrong structure for it.
What makes an outsourced finance relationship work
The most common reason an outsourced finance relationship underperforms has nothing to do with the firm's competence. It is information flow.
Getting timely information out of the business is consistently the hardest part of this work. Not because clients are difficult, but because they are busy running companies, and the request for a signed contract or an explanation of an unusual deposit lands somewhere behind 11 more urgent things. Every week that goes by, the close slips and the numbers get less useful.
So the minimum requirement is one engaged point of contact who can be reached, answer questions, and make decisions. Not a lot of hours. Reachable. Where that exists, these relationships tend to last for years. Where it does not, no amount of expertise on our side compensates for it.
The outcome we are actually aiming at
There is a version of success in this work that sounds strange coming from an outsourced firm: the business ends up able to bring the function in-house if it ever wants to.
That is not the outcome we push for, and most companies never take it. The economics of a full finance department rarely make sense at this size, which is the whole reason this arrangement exists. But it is the standard the work is held to. A finance function that is genuinely running, with a close that holds and reporting somebody could pick up on a Monday, is one a company could take back if it chose to. One that only works because we are the ones touching it is not.
There are a great many firms that will keep your books. Fewer are built so that a client could leave.
Common questions about bookkeepers, controllers, and CFOs
What is the difference between this and a CPA firm?
A CPA firm handles compliance work: audits, filings, year-end tax obligations. We handle the ongoing financial operations of the business: the day-to-day accounting, the strategic financial picture, and the systems that make everything run.
We work alongside your tax CPA rather than instead of them. That coordination is a service in itself, which we call CPA support. They do their job, we do ours, and audit season is considerably less painful because the books and documentation are ready when they ask.
How is this different from hiring a bookkeeper?
A bookkeeper records what happened. We help you understand what it means and what to do about it. Depending on what the business needs, that is controller-level work, fractional CFO support, or both, and it changes over time as the company does.
Do I need all three roles at once?
Almost never. These are three different jobs, not three sizes of the same one. Recording is the layer nearly every business already has, so the real question is which of the two above it is thin: verification, or judgment.
They stack rather than replace each other. Adding a CFO does not remove the need for someone verifying the numbers underneath, which is why the answer is rarely to swap one for another.
Do we have to change accounting systems to work with you?
No. We meet businesses in the systems they already have, and most engagements start exactly that way. If we think the software has become the limiting factor, we will say so and explain why, but that is a recommendation with reasoning behind it, not a precondition for working together.
How does pricing work?
It depends on scope: which functions need covering and how much involvement makes sense. Pricing is based on the work rather than an hourly rate, so the conversation starts with what the business actually needs rather than with a rate card.
Do you work with businesses virtually?
Yes. Most of the work is done virtually. Our focus is Georgia, Florida, and Tennessee, and we work across the broader Southeast and nationally where the fit is right.
What size company is this built for?
Typically businesses between $2 million and $50 million in revenue. The point where a second and third layer earn their keep usually arrives somewhere past $5 million, and by $25 million it is rarely a question. Complexity has arrived, but a full in-house finance department still does not make sense.
Not sure which of these you are
That is a reasonable place to be, and it is a short conversation rather than a long one. Tell us where you are trying to take the business and what your month-end looks like now, and APN Group will give you an honest read on what the next layer should be, including if the answer is that you are fine where you are.