Anonymised, illustrative composite. Every bid the company had ever won went through one person's head and nobody else's — and no buyer was prepared to pay full price for a business that could not price its own next job without him.
At a glance
A trades contractor with reported seller's discretionary earnings of $500,000 came to market at a valuation built from deavo's published trades-sector median SDE multiple of 2.9 times — a $1,450,000 baseline. See deavo's valuation guidance for the sector medians the buyer's advisors started from.
Diligence surfaced a single-person dependency that the earnings multiple alone did not capture: every bid the company had ever submitted was prepared personally by the founder, using judgment and relationships built up over two decades, with no written estimating process, no second estimator, and no documentation of how a quote was actually built. Deavo's own published guidance on this exact pattern says that “brokers and buyers commonly describe owner-dependence as a discount applied to otherwise comparable businesses,” while deliberately declining to size it: see deavo's owner-dependence guidance, which states plainly that “the size of the effect depends heavily on the specific industry and how replaceable the owner's role actually is” and offers no number to attach to it.
Rather than negotiate a discount off the $1,450,000 baseline — a number neither side could actually support with a documented percentage — the parties restructured the payment timing instead. $1,015,000, 70% of the baseline value, was paid at closing. The remaining $435,000, 30%, was structured as an earn-out contingent on the founder training and successfully handing estimating responsibility to a second, buyer-hired estimator over an 18-month transition period.
Deavo's guidance names this restructuring pattern directly — extended transition periods, earn-out structures and non-compete agreements, rather than a straight price cut, as the standard market response to owner dependence. The tax mechanics are the part most often assumed rather than checked, and the assumption is wrong here. The capital gains reserve in subparagraph 40(1)(a)(iii) of the Income Tax Act is available only for “such of the proceeds of disposition of the property that are payable to the taxpayer after the end of the year” — clause (C) — and its size is then capped by clause (D) at a fraction that forces at least one-fifth of the gain into income each year, so a reserve can run for at most five taxation years including the year of disposition. A milestone-contingent earn-out is not payable: until the second estimator was hired and trained, no amount was owing at all. So the reserve did nothing for the 30% here — the deferral came from the contingency itself, not from s. 40(1)(a)(iii), and how a contingent earn-out is ultimately reported is a matter of CRA administrative practice rather than of this subparagraph. The ten-year reserve would not have helped either: it lives in subsections 40(1.1) to (1.4) and reaches only dispositions to a child, qualifying intergenerational business transfers, employee ownership trusts and worker cooperatives — none of them an arm's-length sale to an outside buyer.
The founder stayed on through the transition period as agreed, the second estimator was hired and trained inside the 18-month window, and the earn-out paid out in full once the milestones were met. The buyer got a business that no longer depended on one person's judgment to price its next job; the founder captured the full baseline value rather than a discounted price with no path back to it. For the related file where a lender, not a buyer, discounted a concentrated risk directly in its underwriting, see sixty per cent of revenue from one customer, and for the mechanics of the deferred-payment structure itself, the earn-out glossary entry.
A straight price cut negotiated purely on owner-dependence risk, with no published figure either side could point to, would likely have landed lower than the earn-out's effective value and left the founder with no way to recapture it even after successfully training a replacement. Structuring the risk into a contingent, milestone-based earn-out instead let the price reflect what the business was worth once the dependency was actually fixed, not what it was worth on day one with the risk still live — and it gave the buyer real protection if the transition failed, rather than a price cut it had already paid for regardless of the outcome.
The tell was structural, not financial: a business with no documented estimating process and exactly one person who could produce a bid. A valuation built on a clean earnings multiple will not surface that risk on its own — it shows up only when diligence asks who actually does the work behind the numbers, and what happens to the pipeline the day that person stops showing up.
A 30-minute call is enough to tell you whether AI pays for itself here.
