Anonymised, illustrative composite. The technicians had done the work. The buyer’s accountant just couldn’t find it anywhere the money was supposed to be.
At a glance
A buyer was acquiring an Ontario HVAC and mechanical-service business built on annual maintenance-plan customers — the recurring-revenue base deavo’s Canadian trades guide identifies as the thing that actually moves a multiple: “Recurring service contracts, crew retention and fleet condition move the multiple more than last year’s revenue.” The seller reported trailing owner-discretionary earnings of $510,000. The letter of intent priced the deal at 3.1× SDE — $1,581,000 — inside deavo’s published trades benchmark of 2.5–4.0×, a range the page states are “illustrative ranges based on comparable Canadian transactions, not a valuation, deal or investment opinion.”
A field-service software migration mid-year had left a gap: technicians had completed roughly two months of commercial maintenance visits that were never entered into the new invoicing system before the old one was decommissioned. The buyer’s accountant found the work in job-completion logs, not in accounts receivable. It had never been billed, and it was not reflected anywhere in the seller’s closing net working capital statement. It also carried a second, quieter exposure: under Excise Tax Act s.152(1)(b), consideration for a taxable supply is deemed to become due on “the day the supplier would have, but for an undue delay, issued an invoice” — so the GST/HST on completed work does not wait for the invoice that a software migration swallowed. The parties confirmed with the seller’s accountant that the affected periods were re-filed before the true-up ran.
The unbilled work totalled $85,000 across the affected customer accounts. Rather than dispute the SDE figure or the multiple, the buyer treated it as a working-capital peg problem: the $85,000 was excluded from the agreed net working capital target because it did not appear in the accounting records the peg was set against, so it had to be resolved through a post-closing true-up instead of a price renegotiation. The full $85,000 was held back in escrow at closing, dropping cash to the seller at signing to $1,496,000 ($1,581,000 − $85,000). Over a 90-day collection window, the buyer’s bookkeeper invoiced every affected customer: $71,000 was collected (83.5% of the held-back amount), and $14,000 was written off against customers who had already cancelled their maintenance plans before the gap was found. The escrow released $71,000 to the seller and returned $14,000 to the buyer as a net working capital adjustment, for a final adjusted price of $1,567,000 ($1,581,000 − $14,000).
An escrow tied to a specific, identified working-capital gap is a market mechanic, not a statute, but Ontario practice treats it consistently: as treadstonelaw’s own guidance puts it, funds are “retained, either in a lawyer’s trust account or with an escrow agent, for a set period so the buyer has a source of payment if the seller’s representations turn out to be inaccurate or a pre-closing liability surfaces” — and because “there is no statutory period,” the 90-day window here was set to match how long collection actually took, not a market convention borrowed from a different kind of deal.
The buyer paid for the recurring maintenance-plan base the way deavo’s own guidance frames it — for contracts that survive the handover — and paid nothing for the $14,000 of work that, once tested, never converted to cash. The seller kept 83.5 cents of every unbilled dollar the buyer could actually collect, without the deal collapsing into a renegotiation of the multiple itself.
Had the buyer accepted the seller’s closing net working capital statement at face value and paid the full $1,581,000 at signing, it would have overpaid by exactly the $14,000 that later proved uncollectible — a small figure against the deal, but one the escrow structure recovered entirely, at zero cost to the seller for the $71,000 that was genuinely owed.
The tell was the software cutover date itself: any transition between field-service or invoicing systems is the single most reliable place unbilled work hides, because completed jobs can sit in the old system’s queue with no bridge into the new one. Reconciling technician job-completion logs against the invoicing system for the sixty days on either side of any such migration is a cheap check that would have surfaced this before the letter of intent was signed, not after.
It is also cheaper than the alternative diligence gap deavo flags for this sector on the same page — owner dependence, its stated “#1 Diligence snag” for trades businesses — because a job-log reconciliation is a mechanical, hours-long exercise, while proving a business runs without its founder standing in the truck bay usually is not. The two risks compound: a firm whose billing runs through one person’s memory of who was on-site is exactly the firm most likely to lose two months of invoices in a system change.
A 30-minute call can tell you whether a working-capital peg on your next deal is set against the right numbers.
