The question usually arrives in a frustrated form. The numbers came late again, or they came on time and nobody in the room believed them, and the conclusion is that the finance setup needs to be more senior. Sometimes that is right. Often the company is about to spend CFO money to fix a bookkeeping problem, which does not work and is expensive to discover.
The distinction is cleaner than the job titles suggest. If you do not trust your numbers, or they arrive weeks late, that is a bookkeeping problem and a CFO will not fix it. If your numbers are clean and current but you cannot answer what happens to runway if you hire six engineers, that is the CFO gap.
The four-question test
Work through these in order. The first one you answer no to tells you what to fix.
- Are your books closed within about a week of month end? If not, you have a bookkeeping problem. Everything downstream is guesswork until this is true.
- Do you trust the numbers enough to send them to an investor without caveats? If not, it is still bookkeeping, usually accrual treatment or revenue recognition.
- Can you answer a scenario question, such as what happens to runway if we hire six people in Q3, in an afternoon? If not, that is the CFO gap.
- Can you explain which customers or products are actually profitable? If not, that is also the CFO gap, and usually the more valuable half of it.
Questions one and two are about the record. Three and four are about what you do with it. No amount of seniority fixes a late ledger, and no bookkeeper, however good, is going to build your hiring scenarios.
What each layer actually is
The roles are not a smooth gradient of seniority, they are different jobs with a real classification boundary between them. The Bureau of Labor Statistics separates them explicitly, and the definitions are more useful than any vendor comparison table.
| Role | What the definition covers | US median pay |
|---|---|---|
| Bookkeeper | Routine calculating, posting, and verifying. Recording, not interpreting | $50,670 |
| Accountant | Prepares financial statements. The word prepare is the dividing line | $83,680 |
| Controller | Classified under financial managers. Owns the close and the accuracy of the output | $166,570 |
| CFO | Classified under chief executives. Owns the forward-looking decisions | $213,990 |
Two things in that table catch people out. Bookkeeping is the only one of the four where preparing financial statements is not part of the definition, which is exactly why a bookkeeper-only setup struggles the moment an investor asks for statements. And CFOs are classified with chief executives rather than financial managers, so the widely quoted $166,570 is the controller tier, not a CFO salary. Most comparison articles get this wrong by about $47,000.
Worth noting on trajectory too: bookkeeping clerk employment is projected to decline 6% through 2034 while financial manager roles grow 15%. The routine recording work is being absorbed by software. The judgment work is not.
The sequencing mistake
The expensive version of this decision goes as follows. Reporting is unreliable, so the company hires a fractional CFO. The CFO opens the books, finds them three weeks behind and partly on a cash basis, and spends the first two months cleaning up. That cleanup is bookkeeping work billed at CFO rates, and at the end of it the company has current books and has not yet received any of the strategic work it was paying for.
The order that works is to get the ledger current and accrual-based first, then add the forward-looking layer on top. If you are buying both anyway, buy them from a team that already does both, so nobody is waiting on a handoff.
Symptom to fix
| What you are experiencing | What actually fixes it |
|---|---|
| Books close three weeks after month end | Bookkeeping: cadence and process |
| Investor asked for GAAP financials, yours are cash-basis | Bookkeeping: accrual conversion |
| Nobody can tell you this month's burn without a spreadsheet | Bookkeeping: current ledger, then a dashboard |
| You cannot model a hiring plan against runway | Fractional CFO |
| You do not know which segment is profitable | Fractional CFO |
| You are raising in six months and the model worries you | Fractional CFO |
When you need both
Most funded companies do, and the trigger is usually a raise. Investors read the statements and ask questions the model has to answer, so the record and the interpretation both have to hold up in the same conversation. For the layered version of this question, including where a controller fits, see bookkeeper vs accountant vs fractional CFO.
One useful mechanical trigger for the next decision along: when your outsourced finance spend passes roughly $33,000 a month, building the function in-house generally starts to make economic sense. Below that, buying it is cheaper than hiring it.
Where Zinance fits
Zinance puts both layers on one team. The books close daily, so the record is never the bottleneck, and the fractional CFO work runs off numbers that are already current rather than waiting on a close. You are not managing a handoff between a bookkeeper and an advisor who disagree about which figure is right.