The same post-closing decision produces a completely different earn-out outcome depending on one choice made months earlier at signing: which metric the payment actually tracks. Get the metric wrong and every other protection in the agreement is defending the wrong number.
Key takeaways
Treadstonelaw.ca’s own definition of an earn-out identifies what typically triggers the payment: “post-closing business performance (revenue, profit, or similar metrics).” That short list hides a real decision. Each of those options protects a different party against a different risk, and none of them is free of a weakness.
Top-line revenue sits above the line where most of the accounting disputes in this cluster actually happen — receivable collectibility, inventory valuation, cost allocation, one-time-item treatment. A buyer has fewer levers to move revenue after closing than to move profit, which makes it the metric least exposed to manipulation, deliberate or otherwise. The cost is that revenue says almost nothing about whether the business is actually healthy: a buyer could hit a stated revenue target by discounting margin toward zero, and the earn-out would pay out in full on a business quietly being run at a loss.
Gross profit sits closer to the business’s real economics than revenue does, factoring in the direct cost of what was sold. It is correspondingly more exposed than revenue to exactly the kind of dispute that recurs in this cluster — how cost of goods sold is allocated, which costs are treated as direct versus indirect — without yet capturing the operating-expense decisions further down the income statement.
An earnings-based metric like EBITDA reflects the business’s overall health most completely — and is exactly the number a buyer’s post-close decisions move the most. A treadstonelaw.ca article on earn-out disputes names the pattern directly: the buyer “changes pricing, product lines, or service offerings in ways that depress the metrics the earn-out is measured against.” A companion piece is blunter still: earnouts “produce the ugliest fights, and almost always about how the business was run after closing rather than about arithmetic.” EBITDA gives the fullest picture of business health precisely because it sits downstream of every decision a buyer now controls — staffing, marketing spend, product mix — which is also exactly why it is the metric most exposed to a dispute.
Some agreements measure something more insulated from ordinary operating discretion than any financial figure — a specific customer-retention count, the renewal of a named contract, a defined production volume. This trades comprehensiveness for resistance to gameability: it is harder for a buyer to move a single named contract’s renewal through general operating decisions than to move EBITDA. The tradeoff is that an operational metric needs its own precise, unambiguous definition, or it inherits the same drafting risk described below rather than avoiding it.
A well-chosen metric with a badly defined measurement period still produces a fight. The same treadstonelaw.ca source names “ambiguity in how the purchase agreement actually defines the target or the measurement period” as its own, independent driver of earn-out disputes — distinct from which metric was chosen in the first place. Precisely stating the start date, end date, and what happens if the business is sold again or restructured before the period ends is not optional drafting; it is where a well-chosen metric can still be undone.
Matching the metric to what you actually want to protect
Choosing revenue over EBITDA, or vice versa, does not settle the accounting question underneath either one. Whichever metric is chosen, the agreement still needs to specify which accounting policies define it and which prevails if they conflict with GAAP — a genuinely separate drafting task covered in full in accounting rules that govern the earn-out calculation. A carefully chosen metric, paired with a vaguely specified accounting basis, buys none of the protection the metric choice was meant to provide.
A worked contrast. Six months after closing, a buyer discontinues a lower-margin product line that had been generating roughly $400,000 of annual revenue at close to break-even. Measured on a revenue-based earn-out, the discontinuation shows up as a straightforward $400,000 shortfall against target — a clear, disputable loss to the seller. Measured on an EBITDA-based earn-out over the same period, the same discontinuation can show a smaller decline, or even a modest improvement, because the discontinued line was margin-dilutive and its removal freed up cost and management attention that improved the profitability of what remained. The same real operating event produces opposite-signed earn-out outcomes depending entirely on which metric the agreement chose — which is exactly why the choice has to be made deliberately, with a specific risk in mind, rather than defaulted to whichever term sheet template was used last.
Not necessarily — it protects against deliberate manipulation, but not against a buyer simply failing to grow the business as fast as the seller expected. A seller confident in the business’s margin resilience may reasonably prefer EBITDA precisely because it captures more of the real outcome.
That is a matter for the agreement’s own drafting, not addressed by any source relied on here — treat it as a negotiation point to raise explicitly rather than an implied entitlement.
Not inherently, but it does move the fight toward exactly the decisions — pricing, staffing, product mix — a buyer controls after closing, per the pattern described in treadstonelaw.ca’s own account of earn-out disputes.
A short call walks through which metric actually protects what you’re trying to protect on your specific deal.
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.
