You are hiring for five roles, not one job
A complete finance function breaks into five distinct roles: a bookkeeper who records, a senior accountant who closes, a controller who reviews and controls, a CFO who frames decisions, and a tax CPA who plans. Job titles blur in owner-managed businesses, and one person often wears two or three of these hats, but the work itself does not blur. When one of the five is missing, its work either lands on you, lands on the wrong person, or quietly does not happen.
This is worth naming before you write a job posting, because the most common posting at this size is a single ad titled "full-cycle accountant" that describes three of the five roles at once. The candidates who can genuinely do all three are rare, expensive and hard to keep, and the candidates you can afford will do one of the three well. Deciding which roles deserve an employee, which should be fractional and which stay external is the actual hiring decision; the individual hires come after.
The same logic applies whether the trigger is growth, a departing bookkeeper, or an accountant who only appears at year-end. You are not replacing a person. You are staffing a function.
And functions are judged on outputs: a close that lands on time, filings that never slip, reports somebody actually reads, and a tax position that was decided rather than discovered. Hold every staffing option, employee, fractional or outside team, to that same output list and the comparison gets much easier.
What each role actually owns
Each role owns a different altitude of the same numbers, and a missing altitude shows up as a specific kind of pain. Bookkeeping gaps look like messy records, unbilled work and vendors calling about late payments. Controller gaps look like reports nobody quite trusts and a close that lands weeks late. CFO gaps look like financing, pricing and expansion decisions made on instinct because nobody framed the numbers for a decision.
| Role | What they own | When a full-time seat makes sense |
|---|---|---|
| Bookkeeper | Daily transaction entry, bank reconciliations, receivables and payables, receipts and supporting records | Driven by transaction volume; usually the first finance hire |
| Senior accountant | Month-end close, working papers, HST returns, payroll runs and remittances | When the close keeps slipping or the bookkeeper is past capacity |
| Controller | Reviewing the close, internal controls, cash forecasting, systems, managing the people above | Typically once finance is two or three people, or reporting drives real management decisions |
| CFO | Financing, pricing, forecasts, corporate structure, lender and investor conversations | Rarely full-time below large-company scale; usually a fractional arrangement |
| Tax CPA | Corporate tax planning, owner compensation, reorganizations, CRA correspondence and reviews | Almost never in-house at this size; bought as external professional expertise |
We will not quote salaries here, because they move with the market and the year and any figure would be stale by the time you read it. The structural point holds regardless: the higher roles cost more per hour precisely because you need fewer of their hours, which is why the top of the stack is normally bought in fractions rather than employed. What matters is that every row in that table has a named owner, even if several rows share one name.
There is also a sensible build order. Volume justifies the bookkeeper first, reporting pressure justifies the senior accountant or controller second, and the CFO and tax layers are bought in fractions from the start, because their value arrives in decisions rather than in hours. Businesses that hire top-down, a controller with nobody to control, or bottom-up only, three bookkeepers and no reviewer, end up paying for the missing altitude either way, in errors or in missed decisions.
Why one great hire cannot be the whole function
One person cannot safely hold the entire function, because basic segregation of duties requires that the hands on the books are not the only eyes on the bank. The person who enters bills should not also approve the payments and reconcile the account with nobody reviewing any of it. That is not an accusation against any employee; it is the control that protects honest employees from suspicion and protects the business from the rare dishonest one.
Errors are the more common cost of a one-person department. Unreviewed work drifts, and a solo accountant marking their own homework has no mechanism for catching a systematic mistake before it reaches an HST return, a T4 run or a lender's covenant calculation. A second layer of review is what turns recorded transactions into numbers you can act on.
The workable minimum is two sets of hands with the duties deliberately split: one person processes transactions and prepares the payment run, a different person approves payments and reviews the reconciliations, and the owner keeps sight of the bank. That can be two employees, or one employee plus an outside team acting as the review layer. What it cannot be is one person and a promise.
Coverage is the third problem. A one-person department takes vacation, gets sick and eventually resigns, and each of those events pauses the function entirely. We wrote separately about what happens when your only finance employee leaves; the short version is that the exposure is far larger than a recruiting fee. Any in-house plan needs at least two people who can each run payroll and the close, or an outside layer that can step in.
Payroll and compliance have to sit with someone
Payroll and compliance form a role in miniature, and it needs a named owner from the first pay run. Source deductions, HST filings, T4 and T5 slips, Employer Health Tax where payroll is large enough, and WSIB where the industry requires it all run on fixed deadlines with automatic penalties for missing them. None of this is intellectually hard. All of it is unforgiving.
Directors can be held personally liable for source deductions that are withheld and not remitted, which makes this the one corner of the function an owner should never leave ambiguous. In small teams the compliance calendar usually defaults to the bookkeeper, and that works while the bookkeeper is present, current and correct. The risk is when it defaults to nobody, which is exactly what happens in the gap between hires or during a leave.
Whoever designs your in-house function should put the owner and the backup for every recurring filing in writing. If you cannot name the backup, you have found the first hole in the plan.
Software helps here and does not solve it. Payroll platforms calculate and remit competently, but they execute what someone tells them; they do not notice a new hire set up in the wrong province, a taxable benefit nobody entered, or a filing frequency CRA has changed. The calendar still needs a human owner who understands what the software is doing on the company's behalf.
The facts that change how many seats you fill
Six facts decide how much of the stack you should actually employ, and they matter more than any org chart copied from a bigger company:
- Transaction volume and entity count. More invoices, locations and corporations mean more bookkeeping hours before anything else grows, and intercompany activity multiplies the close.
- Who consumes the reporting. A lender with covenants, a partner, or managers running budgets pull you up the stack toward controller-level output; if the reports are for you alone, the bar is lower.
- How often real decisions arrive. Financing rounds, pricing changes, expansion and acquisition questions are CFO work whether or not anyone holds the title.
- Your tolerance for being the reviewer. If nobody checks the work, you are the control, which is fine only if you have the time and the training to be one.
- The hiring market you sit in. Experienced controllers are scarce and mobile across the GTA; a plan that depends on hiring one quickly is a plan that depends on luck.
- Your growth rate. A function sized for today is undersized in eighteen months if revenue is compounding, and re-hiring is slower than re-scoping.
Run your own situation against those six honestly. Most businesses in the $750k to $25M range land on one to two employed seats plus fractional and external coverage for the rest, not five hires. The pattern shifts with the shape of the business: a distributor with heavy daily volume staffs differently than a professional practice with forty invoices a month and complex owner tax. The stack is the same; the split is yours.
The other way to staff the stack
The alternative to five hires is to buy the stack as one team, which is how most businesses at this size get all five altitudes covered without carrying five salaries. An Ongoing Financial Partnership runs books, payroll, reporting, tax and advisory as a single outside function, with the bookkeeping, review and tax layers built in and a Fractional CFO layer added where decisions demand it. The segregation of duties and the vacation coverage come with the team instead of being problems you solve twice. The reporting layer scales the same way: month-end lands on a calendar, someone senior reviews it before you see it, and the CFO conversation happens with people who already know the file.
The honest comparison is not salary versus fee; it is the fully loaded cost of the seats you would fill against the scope of the function you would buy, judged on the same output. We laid out what an in-house finance team really costs once you add everything up so you can run that comparison with real numbers. If it would help to talk it through against your own volume and reporting needs, the scoping conversation is a free 15-minute discovery call, and the fee comes back in writing.
