Anonymised, illustrative composite. Nine of a target's twenty-six “contractors” had set hours, a company laptop, and no other clients — and the buyer's diligence team had to price the reclassification risk before signing.
At a glance
A field-services business serving industrial clients reported a workforce of 34 payroll employees and 26 independent contractors on its financial statements, with the contractor arrangement presented in the seller's information memorandum as a deliberate, flexible cost structure that let the business scale technician capacity up and down with demand without payroll overhead.
The buyer's HR diligence workstream sampled a third of the contractor roster against actual working patterns rather than accepting the labels on file. Nine of the nine sampled contractors worked a fixed weekly schedule set by the target's dispatch office, used a company-branded vehicle and company-issued tools, reported daily to a site supervisor, and had performed work exclusively for the target for between two and six years. None had incorporated; several described themselves in interviews as “basically an employee, just paid on invoice.”
Extrapolated across the full 26-person contractor roster using the sample's 100% flag rate as a conservative working estimate, the buyer's tax advisors modelled retroactive exposure of roughly $340,000 — unremitted source deductions, employer CPP and EI contributions, and ESA entitlements including vacation pay these workers had never been credited with, spread across a look-back period the advisors judged realistic given the target's record-keeping. That figure sat entirely outside the seller's reported EBITDA, because a misclassified contractor's true cost as an employee had never been modelled into the business at all.
Why the label matters at all traces back to ITA s.153: “Every person paying…salary, wages or other remuneration…must deduct or withhold from the payment the amount determined in accordance with prescribed rules and must, at the prescribed time, remit that amount to the Receiver General on account of the payee’s tax for the year.” That withholding duty attaches to an employer paying an employee, not to a payer of an independent contractor's invoice — so the question of which one a worker actually is decides who owed the money all along. The test that governs that question is not a checklist, and it does not turn on what either side signed. The Supreme Court of Canada framed the central question in 671122 Ontario Ltd. v. Sagaz Industries Canada Inc. as whether the person engaged to perform the services is performing them as a person in business on their own account, adding that no single test is conclusive and that the level of control “will always be a factor”. The factors weighed under that question — drawn from Wiebe Door Services Ltd. v. M.N.R., which Sagaz adopts — are indicative rather than exhaustive: as one summary of the CRA's own approach puts it, the CRA and the courts weigh control (who decides how, when and where the work happens), tools and equipment, the worker's genuine opportunity for profit and risk of loss, and integration into the business's core operations. Crucially, “mutual agreement on the label is a factor CRA and the courts consider, but it doesn't override the substance of the relationship if the facts point the other way” — a signed contractor agreement, however carefully drafted, does not settle the question on its own. Every signal in this target's sample pointed toward employment: set hours, company tools, a single client, ongoing integration into day-to-day dispatch.
The buyer did not walk, but it did not accept the seller's cost structure at face value either. The $340,000 modelled exposure was built into the indemnity basket as a specifically carved-out, uncapped item rather than folded into the general representations basket, on the reasoning that a known, quantified risk deserves a different treatment than an unknown one. Purchase price was also adjusted downward by roughly the same amount, on the view that the seller's reported margin had been artificially inflated by a cost structure the buyer would have to unwind post-close. Nine of the nine flagged workers were converted to payroll employment within the first sixty days after closing.
The signal to test directly, rather than accepting a contractor count as a cost-structure fact, is whether any given “contractor” works for anyone else. A worker with one client, fixed hours set by that client, and equipment supplied by that client has none of the genuine business risk the classification test looks for — and a business built on treating that arrangement as contractor pay is, in substance, running an unbudgeted payroll liability the seller's financial statements have never had to carry. This sits in the same family of diligence findings as a target's quality-of-earnings report turning up cost-structure adjustments that never showed up in the seller's reported margin.
A 30-minute call is enough to tell you whether AI pays for itself here.
