This decision usually gets made on feel. Outsourcing sounds like a stopgap and hiring sounds like building a real company, so companies hire earlier than the arithmetic supports and then discover that one person cannot cover bookkeeping, tax, payroll and reporting at once.
There is a cleaner way to decide it. Compare what you spend on outsourced finance against the fully loaded cost of the smallest in-house team that would genuinely replace it, and be honest about what that team is.
The comparison people actually make, and why it is wrong
The instinctive comparison is a monthly outsourcing fee against one salary. On those terms outsourcing looks expensive quickly, which is why founders reach for the hire.
The problem is that the two sides are not the same scope. An outsourced engagement typically covers bookkeeping, the close, tax filings, and often credits and reporting. Replacing it in-house is rarely one person. Bookkeeping alone runs to a US median of $50,670, an accountant who can prepare financial statements to $83,680, and a controller who can own the close to $166,570. Add the tax preparer you still have to pay separately, because almost nobody hires that in-house at this stage.
BLS Occupational Employment and Wage Statistics, May 2025. CFO is classified under Chief Executives, not Financial Managers.
The real cost of an in-house hire
Salary is roughly two thirds of the number. The rest arrives as payroll taxes, benefits, software, recruiting and management time, and it is consistent enough to plan with.
| Cost line | Typical addition |
|---|---|
| Base salary | The figure you are comparing |
| Payroll taxes and benefits | Commonly 20–30% on top |
| Software and tooling | Accounting, payroll, expense, close tools |
| Recruiting | Fee or your own time, plus weeks of vacancy |
| Management | Someone senior reviewing work that used to arrive finished |
| Coverage risk | Holiday, sickness, and the day they resign |
That last row is the one that does not appear in a spreadsheet and matters most early. A single in-house person is a single point of failure over your books, and the failure mode is not that they leave, it is that they leave in the month you are raising.
The threshold worth using
A practical trigger used by startup accounting firms is to build in-house once outsourced finance spend passes roughly $33,000 a month, which is about $400,000 a year. Below that, buying the function is usually cheaper than assembling it.
For the senior layer specifically, a widely used investor heuristic is the same shape: hire the full-time person once you are spending about three quarters of a full-time salary on fractional support. Both rules say the same thing. Outsource until the spend approaches the cost of the team you would build, then build it.
Run the comparison against the team that would actually replace your provider, not against one hire. If your engagement covers books, close, tax and credits, the honest in-house equivalent is usually a bookkeeper plus a controller plus an external tax preparer, and that is a very different number from one salary.
Running the comparison on your own numbers
A threshold is a shortcut. Your own arithmetic is better, and it takes about fifteen minutes. Work out the smallest in-house team that would genuinely replace what you buy today, then cost it fully rather than by salary.
Two companies illustrate why the answer varies more than the threshold suggests.
| Company A | Company B | |
|---|---|---|
| Outsourced spend today | $3,400/mo | $3,400/mo |
| What it covers | Books, tax, R&D, light CFO | Books, tax, R&D, light CFO |
| Transaction volume | Low, one entity, simple revenue | High, three entities, usage billing |
| Smallest replacement | One part-time bookkeeper | One full-time accountant plus external tax |
| Fully loaded replacement cost | About $2,600/mo | About $9,500/mo |
| Which is cheaper | In-house, marginally | Outsourced, substantially |
Same spend, opposite answers. The variable is not size in revenue terms but the complexity of what has to be done each month, because that determines how much person you need rather than how many transactions there are. This is why revenue-based rules of thumb mislead: a $6 million single-product business and a $6 million multi-entity business need very different finance functions.
When you cost the replacement, include employer taxes and benefits, software licences that the provider currently supplies, recruitment cost amortised over expected tenure, and the management time of whoever the hire reports to. That last one is genuinely a cost and is almost always left out, because it lands on a founder who is not counting their own hours.
What actually breaks the build case
Beyond arithmetic, four things tend to make an in-house hire underperform the model, and they are worth knowing because none of them show up in a cost comparison.
- Single-threading. One person means no coverage for holiday, illness, or departure. During a fundraise or an audit, a single point of failure in finance is a genuine risk rather than an inconvenience.
- Seniority mismatch. The work spans routine transaction processing and board-level analysis. One hire is either overqualified for most of the week or underqualified for the important part, and both are expensive in different ways.
- No specialist depth. R&D credits, multi-state tax, and revenue recognition under unusual contracts are periodic specialisms. A generalist hire handles them adequately or brings in help, at which point you are paying for both.
- Ramp time. A new finance hire is not fully productive for months, and that period frequently coincides with whatever prompted the hire, which is usually growth or a raise.
The counterweight is real and worth stating plainly: an in-house person builds context that no external provider matches. They sit in the room, hear the strategy discussion, and know why last quarter looked odd. When the finance work is genuinely strategic rather than procedural, that context is the whole value, and the cost comparison becomes secondary.
What outsourcing genuinely gives up
Being fair about the trade matters, because there are real reasons to build in-house before the arithmetic says so.
- Context. An in-house person absorbs how your business works in a way a provider covering many companies does not, and that shows up in the quality of questions they ask.
- Availability. Someone in your standups and on your Slack all day catches things earlier, though a provider with a fast response time closes much of this gap.
- Bespoke work. Unusual reporting, deep operational analysis, and custom finance projects fit an employee better than a scoped engagement.
- Control over process and systems, if you have strong opinions about how the close should run.
What outsourcing gives you in exchange is a team rather than a person, coverage that does not depend on one individual, and a cost that scales with your complexity rather than in salary-sized steps.
The sequence most companies end up following
- Pre-seed to seed: outsource everything. The volume does not justify a hire and the work is mostly standardised.
- Post-Series A: outsource the operating layer, add fractional CFO support around the raise and the board.
- Approaching $400,000 a year in outsourced spend: hire a controller or head of finance in-house and keep the specialist work outside.
- Growth: build the team, retaining outside specialists for tax, credits and valuations, which almost nobody brings fully in-house.
Notice that the last step still buys some things. Tax and R&D credits stay outsourced at almost every size, because the expertise is episodic and expensive to keep on staff.
Where Zinance fits
Zinance is built for the stretch before an in-house team makes sense, and to hand over cleanly when it does. Books close daily, a dedicated team answers on Slack in about ten minutes, and bookkeeping, tax and R&D credits sit in one engagement from $349 a month. Because the ledger lives in your own QuickBooks file, moving the operating layer in-house later is a handover rather than a migration. For the role-by-role version of this question, see bookkeeper vs accountant vs fractional CFO.
If the answer is to keep buying it for now, the next question is what to buy and from whom: the best bookkeeping services for VC-backed startups compares the market on the things that actually differ. If you already have a provider you have outgrown, how to switch without dropping a month covers the handover.