The question usually surfaces about six months before a round, often prompted by an investor asking who owns finance. The instinct is to hire, and it is almost always the wrong first move. A full-time CFO at this stage is one of the most expensive hires you can make, and the work you actually need is a fraction of a full-time role.
You do not need a full-time CFO to raise a Series A. What you need is diligence-ready books, a model that survives investor questions, and someone who can defend the numbers in the room. That is a fractional CFO's job, and the useful window is roughly three to six months before you open the round.
What the guidance actually says, and what it does not
This is a topic where confident numbers circulate without evidence behind them. No independent survey measures when startups hire a fractional CFO, so treat any precise threshold you read as opinion rather than data.
The most substantive datapoint available comes from Bessemer, where 37% of respondents pointed at $10 million to $25 million in ARR as the right point for a first full-time CFO. Investor guidance elsewhere disagrees on where the full-time handoff belongs, with views ranging from Series B to considerably later. What the sources do converge on is the sequence: fractional first, full-time later, with the fractional engagement often beginning around three months before a fundraise.
There is a genuine dissenting view worth knowing. A16z has argued publicly against outsourcing finance at all, on the grounds that it reduces the function to low-level accounting and never advances the business. That position dates from 2017 and predates the current fractional market, but it is a real argument rather than a strawman, and it points at something true: a fractional engagement that never gets beyond bookkeeping oversight is not delivering what you are paying for.
What investors actually check
Series A diligence is a review exercise rather than an audit, and it is mostly about whether your numbers hold together when someone asks where they came from.
- Accrual, GAAP-ready financials with a consistent revenue recognition policy applied all year, not converted the month before the round.
- A model that reconciles to your actuals rather than diverging from them, with assumptions you can defend line by line.
- Metrics that tie back to the statements. If the ARR in your deck cannot be reconciled to revenue in your P&L, the deck stops being evidence.
- A clean cap table, with SAFEs, options, and any 409A work reflected properly.
- Consistency between what you told existing investors all year and what you are telling a new one now.
That last one is where fractional CFOs earn their fee most reliably. A restatement discovered mid-diligence costs more time than a year of proper reporting would have. The Series A audit-readiness checklist covers the full list.
The cost comparison, honestly
One 2026 survey of 346 CFOs puts total compensation including equity at $447,600, with median equity value of $1.5 million. Federal data classifies CFOs under chief executives at a national median of $213,990. Published fractional retainers from named startup providers start around $1,600 a month, which is roughly $19,000 to $50,000 a year depending on tier.
The gap is wide enough that the question is not really cost, it is whether the work needs a full-time person. Before a Series A it usually does not. A useful rule for later: when you are spending about three quarters of a full-time salary on fractional support, make the hire.
The three to six month runway
| When | What should be happening |
|---|---|
| 6 months out | Books current and on accrual; revenue policy settled and applied consistently |
| 4 months out | Model built and reconciling to actuals; metric definitions agreed |
| 3 months out | Data room assembled; diligence question list worked through |
| 1 month out | Numbers rehearsed; the awkward questions answered before an investor asks them |
The sequence matters more than the timeline. Starting at month three with books that are three weeks behind means spending the first month on cleanup, which is the same expensive mistake as hiring a CFO to fix bookkeeping.
The first sixty days
An engagement that starts well looks similar across companies, because the early work is diagnostic rather than strategic. If a fractional CFO proposes to begin by building your forecast, they are starting in the wrong place, and the forecast will inherit whatever is wrong underneath it.
- Weeks one and two: establish what is true. Read the last twelve months of statements, find where the accounting basis changes or the treatment shifts, and identify which numbers can be relied on. This produces an unglamorous list of problems, and that list is the actual deliverable of the first fortnight.
- Weeks three and four: fix the definitions. Agree what ARR means, what sits in cost of revenue, whether burn is gross or net. Write them down. Everything downstream inherits these, so leaving them implicit means rebuilding later.
- Weeks five and six: build the model against actuals. Not a fresh forecast, but one anchored to the closed months, so the first thing it demonstrates is that it reproduces reality before it predicts anything.
- Weeks seven and eight: stress-test it. What breaks the plan, which two assumptions matter most, what the downside case actually looks like. This is what you will be asked in the room, and it is the point of the whole exercise.
If eight weeks in you have current books, written definitions, a model that reconciles, and a defensible downside case, the engagement is working. If you have a beautiful forecast built on numbers nobody has verified, it is not, and the gap will surface during diligence rather than before it.
How to scope the engagement
Fractional arrangements are sold in several shapes and the shape matters more than the rate. Three questions settle most of it.
- Retainer or project? A raise is a project with an end, and pricing it as one is reasonable. Ongoing board reporting is a retainer. Buying a retainer when you need a project, or the reverse, is where most dissatisfaction comes from.
- What happens in a heavy month? Diligence is not evenly distributed. Ask explicitly what happens when the workload spikes, because a strict hourly cap means the arrangement is least available exactly when it matters most.
- Who does the work? In some firms the person who sold the engagement is not the person who delivers it. Ask who will actually be in your weekly call, and what else they carry.
On notice periods, keep them short in both directions early on. A fractional CFO relationship is a judgement fit as much as a skills fit, and finding out it is wrong is much cheaper in month two than in month six of a twelve-month commitment.
Signs you left it too late
The window is three to six months out. These are the symptoms of starting inside it, and each one converts preparation time into remediation time.
- Your first conversation with the CFO is about converting to accrual, rather than about the model.
- You are assembling the data room and the model at the same time, which means neither gets checked against the other.
- A metric in your deck cannot be reconciled to the P&L, and the fix requires restating months already reported to existing investors.
- You are answering diligence questions and building the answers simultaneously, which is how inconsistencies enter the record.
When you genuinely do not need one yet
If you are pre-seed, pre-revenue, and not raising in the next year, a fractional CFO is premature. What you need is a current ledger, accrual from the start so you never have to convert, and someone competent handling tax and compliance. Add the forward-looking layer when there is a decision that needs it.
Where Zinance fits
Because the books close daily, the diligence-ready part is already true rather than something to build before a raise. Bookkeeping, tax, R&D credits, and fractional CFO support sit with one team, so the model, the statements, and the return are all built from one set of numbers when an investor starts checking.