The moment arrives in most growing practices: there is more work than the current team can absorb, and something has to change. The default answer is to hire. Sometimes that is right. Often it is expensive muscle memory.
Here is a way to think it through.
Start with what hiring actually costs
Firms tend to compare an outsourced rate against a salary, which understates the hire considerably. The real comparison includes:
- Recruiting time. Weeks of screening, interviewing and following up, most of it done by someone billable.
- The ramp. A competent bookkeeper is not productive on your files on day one. Three months to full contribution is a reasonable planning assumption, and longer if your files are complex.
- Supervision. Somebody senior reviews the new person’s work closely for the first while. That time is real and it comes out of your capacity, not theirs.
- Fixed cost through slow periods. The salary does not shrink in your quiet quarter. Employer contributions, benefits, software seats and workspace do not either.
- Turnover risk. If it does not work out — or they leave in eighteen months — you pay the recruiting and ramp cost again, and the file knowledge goes with them.
None of this argues against hiring. It argues for comparing the whole cost rather than the headline number.
Then ask what kind of capacity you need
The two options solve genuinely different problems.
Hiring makes sense when:
- The work requires being physically present, or deeply embedded in client conversations
- You need judgment and client-facing capability, not processing throughput
- The volume is stable and predictable enough to justify a fixed cost
- You are building a team you intend to develop into seniors and eventually partners
- You have the supervisory capacity to train someone properly right now
Outsourcing makes sense when:
- The bottleneck is processing volume — entry, coding, reconciliation
- Demand is uneven across the year and a fixed cost would sit idle
- You need capacity in weeks, not months
- You want to test whether the extra clients are actually profitable before committing to payroll
- You do not currently have the senior time to train and supervise a new hire
Read those honestly. Most firms in a capacity crunch are describing the second list and reaching for the first solution.
The question that usually settles it
What is the marginal hour going into?
Track a normal week and split your team’s bookkeeping hours into two buckets:
- Processing — entering bills, coding transactions, chasing receipts, reconciling accounts, hunting unmatched items
- Judgment — reviewing, advising, deciding, explaining things to clients
If the ratio is heavily weighted to the first bucket, you have a throughput problem, and outsourcing addresses it directly. If a meaningful share of the hours is in the second bucket, you need a person — because judgment does not delegate to a process.
Firms are often surprised by the ratio when they measure it rather than estimate it.
The option most people skip
These are not mutually exclusive, and treating them as a binary is the most common error in this decision.
A common sequence that works well:
- Outsource the processing to free up existing senior capacity immediately
- Take on the additional clients the freed capacity allows
- Hire once the revenue is proven — and hire for the client-facing, judgment-heavy role rather than for data entry
This sequence has a useful property: it de-risks the hire. You are no longer hiring on a forecast. You are hiring against revenue that already exists, into a role that is more senior, more interesting to a good candidate, and harder to outsource.
It also inverts the usual failure mode, where a firm hires a junior for processing work, then finds itself unable to move that person up because there is nobody to take the processing.
A note on control
The objection to outsourcing is usually framed as control, and it is worth examining, because control is not actually what most firms mean.
What they mean is predictability: work arriving on time, to a known standard, without surprises. An employee sitting in your office does not automatically deliver that — plenty of firms have had exactly the opposite experience with a hire.
Predictability comes from documented procedures, defined delivery dates, and a review step before work reaches you. Those are properties of an arrangement, not of an employment relationship. Ask any prospective provider how they deliver each one, and judge the answer on the specifics.
The test that costs almost nothing
Whichever way you are leaning, there is an experiment available that a hiring decision does not offer: move one file, for one period, to an outsourced provider.
You will learn what the close package looks like, how the questions arrive, how much of your review time it actually consumes, and whether the arrangement fits the way your firm works. The cost is one month on one file.
Compare that to the cost of learning the same things three months into a hire, and it is not a close call. Run the experiment first, then decide.
apar handles full-cycle bookkeeping behind Canadian accounting firms — starting with one file, with no minimum commitment. See how the first month works.