Most buy-and-build failures are not bad individual acquisitions. They are risks that only exist because several deals were stitched together, and no one checked the group-level effect.
Key takeaways
A fund that has closed three or four add-ons without incident can still be carrying risk that never showed up in any single deal's diligence, because it only exists at the group level. The recurring failure modes in Canadian buy-and-build tend to share that shape: each one is invisible from inside a single acquisition and only becomes real once several businesses are operating under one ownership structure.
This is the risk most likely to be underweighted precisely because it feels like ordinary integration housekeeping. Competition Act s.45(1.1) makes it an offence for “every person who is an employer” who, with “another employer who is not affiliated,” agrees “to fix, maintain, decrease or control salaries, wages or terms and conditions of employment,” or “to not solicit or hire each other’s employees.” The penalty is indictable — “imprisonment for a term not exceeding 14 years… or both” — and s.45(3) lets a court infer the agreement from circumstantial evidence. Every buy-and-build programme, every non-solicit handshake between a platform and a competitor it has not yet acquired, and every informal “we won’t poach each other’s technicians” understanding sits directly against this section. The ancillary restraints defence in s.45(4) only protects a restraint that is “ancillary to a broader or separate agreement” and “directly related to and reasonably necessary” for it — which is exactly why a non-solicit clause belongs inside a signed purchase agreement, and not in a conversation that happens before one exists. See sequencing add-on acquisitions over three years for how this constrains deal timing.
Ontario’s severance pay obligation sits on top of statutory notice, and it is where a roll-up’s scale can quietly change an outcome. An employee qualifies for severance where they have worked five or more years and the employer “has a global payroll of at least $2.5 million,” with the maximum payable capped at 26 weeks. “Global payroll” means exactly that — the whole employer’s payroll, not any one location’s or any one acquired subsidiary’s. A sub-scale add-on that would never have crossed $2.5 million in payroll on its own can sit inside a group that does, the moment the employing entity is the platform rather than the original small business. Reviewing each add-on’s employment exposure in isolation, deal by deal, misses exactly this: the trigger is a property of the group, and it can appear the day an acquisition closes even though nothing changed inside the add-on itself.
The same continuity rule works in the platform’s favour on notice, and against a piecemeal diligence approach if it is ignored. Ontario’s continuity provisions mean a seller’s employees’ length of service “flows through” to the purchaser rather than resetting — the guide’s own example is an employee with ten years at the seller who, terminated a year after the transfer, “will be entitled to eight weeks’ notice rather than just one week.” Each add-on therefore arrives with accrued notice liability the purchase price should already reflect, on top of whatever new severance exposure the group’s combined payroll creates.
Ontario has banned employee non-competes since October 25, 2021, with a narrow sale-of-business exception. Read closely, that exception may not cover the shape most roll-ups actually take: it applies where “there is a sale or lease of a business or a part of a business that is operated as a sole proprietorship or a partnership” and the seller becomes an employee of the purchaser as part of the sale. As written on the regulator’s own page, that names a sole proprietorship or a partnership — not a corporation — and an ordinary Canadian roll-up is almost always buying shares of, or assets from, a corporation. A separate exception exists for named executive roles, and pre-October-2021 agreements remain valid, but a fund that assumes a founder-turned-manager’s non-compete is automatically enforceable because the deal was a “sale of business” is relying on an exception that, on the regulator’s own text, may not name the entity it was acquired from.
The multiple gap that funds a roll-up — see why add-ons price lower than the platform — assumes owner-dependent SDE gets converted into professionalized EBITDA as each add-on integrates. If too many add-ons stay dependent on their original owner, or integration cadence falls behind acquisition pace and no add-on is ever fully absorbed, the platform arrives at exit still looking, to a buyer’s diligence team, like a collection of small owner-run businesses rather than one professionally managed company. Deavo’s own valuation adjusters name “owner-dependence” as a factor that pulls a multiple down — the same factor the platform was supposed to remove is what an exit buyer will price against if it was never actually addressed.
A platform has closed three trades add-ons over two years. Individually, none crossed $1 million in annual payroll, and each was diligenced as a standalone acquisition. Combined, the group’s payroll across the three operating subsidiaries and the head office reaches $2.6 million — over the $2.5 million severance threshold. A technician with six completed years at the first add-on, hired before the platform existed, is let go during an integration restructuring. Statutory notice under the continuity rule reflects the full six years; severance pay is now also owed, calculated as regular weekly wages multiplied by completed years plus completed months divided by twelve — six weeks in this case, well under the 26-week cap, but an obligation that would not have existed had the same employee been let go by the original small business before it was ever acquired. See a bolt-on that broke the platform's service model and three add-ons in eighteen months, one brand for anonymised, illustrative examples of integration failure playing out.
Yes. Section 45(1.1) names “every person who is an employer” with no revenue or headcount threshold anywhere in the provision -- it applies identically to a two-truck landscaping company and a $200 million platform.
The sourced guidance is that the trigger runs on the employer's global payroll, and does not turn on whether operations sit in one legal entity or several under common ownership. Confirm the current interpretation with employment counsel before relying on entity separation to manage this exposure.
No -- it is not banned outright, but the ESA's stated sale-of-business exception names a sole proprietorship or partnership, not a corporation, and a separate executive-role exception may apply instead. Confirm which exception actually fits the deal's structure rather than assuming a share sale is automatically covered.
A 30-minute call is enough to tell you whether AI pays for itself here.
Nobody publishes Canadian transaction data, so every valuation in this country quotes an American benchmark. We are building the Canadian one — multiples, asking-to-sale spreads and days on market, by sector and by city. Leave an email and you will see it first.
No pitch, no listings. One email when the first report lands.
