Adjusted EBITDA diligence questions for home-services roll-ups

1On financed jobs, does every brand’s operating report net dealer fees against revenue, or report them separately, and are finance-company holdbacks and recourse tracked by brand?

In-home window, bath and roofing sales are often financed, and the fee the finance company charges is a real cost of the sale. A brand that nets it against revenue and one that records it in cost of sales show different job values and margins; a brand that books it as interest expense shows higher EBITDA for the same sale. Holdbacks and recourse matter for the same reason: they decide how much of a financed sale the platform actually keeps.

2Are cancellations, three-day rescissions and financing declines removed by the same rule and cutoff at every brand?

In-home sales carry a cancellation right of three business days under the FTC’s Cooling-Off Rule, which some states extend, and financed sales can fail after signing when the customer is declined. A brand that removes these within the month and one that removes them when they are discovered report different revenue for the same period. The reviewer is looking for one rule and one cutoff.

3At every brand, are sales commissions in the same line, expensed in the same period as the job’s revenue, and clawed back on cancellation?

Where a commission sits moves gross margin; when it is expensed moves the period’s EBITDA. A brand that expenses commission at signing and records the revenue at installation can show a weak month followed by a strong one with nothing changed in the business. Clawbacks on cancelled jobs are where the two most often part ways.

4Is each brand’s warranty reserve sized to its own claims history, including lifetime workmanship warranties, and is maintenance-agreement revenue deferred over the plan term?

Lifetime workmanship warranties are common in roofing, windows and bath, and a reserve set before the brand had much claims history may be too small. In HVAC and plumbing, maintenance agreements are paid in advance and earned over the plan term. Both are places where a seller’s books and a buyer’s view of earnings often differ.

5For jobs open at month-end, and for storm work where insurance supplements arrive after signing, does every brand recognize revenue on the same basis?

A roof replaced over two days, or a bath remodel spanning a month-end, has to be recognized somewhere. Storm work adds a second layer: the contract value can rise after signing as the insurer approves supplements. Two brands that handle either differently will not report the same month the same way. Storm-driven revenue in the trailing twelve months is itself a candidate for normalization.

6Before each brand conformed to platform policy, was its history kept on a cash, tax or completed-job basis, how was it restated, and did a system migration change a definition partway through?

Most acquired contractors kept their books for tax and for the owner, not for a lender. The months before a brand joined, and sometimes the months before it moved onto the platform’s systems, sit on a different basis from everything since. A trend line that crosses that boundary without a stated restatement compares two different measurements. Separating EBITDA growth from changes in basis

7For brands acquired mid-period, which pre-acquisition months in the pro forma came from the seller’s records, and which are missing? Are missing months shown as missing, or filled in?

Pro forma EBITDA for brands bought partway through the year is often the largest adjustment in a roll-up. The reviewer wants to know which months were observed and which were not. A month that was never supplied is a gap to disclose; a month filled in by estimate is a figure that needs its own support.

8In each brand’s pre-acquisition books, were deposits on signed but uninstalled jobs ever recorded as revenue, and does each brand’s backlog reconcile to its deposit balance?

Home-services customers often pay a deposit at signing, weeks before installation. Kept on a cash basis, that deposit reads as revenue. Restated, it is a liability until the job is done. The reconciliation between backlog and deposits shows whether the restatement was made, and it feeds the working-capital and net-debt conversation as well as EBITDA.

9Which add-backs apply to each brand, such as owner compensation above a market salary, related-party rent above market, or storm-response costs, and is any recurring cost presented as one-time?

Every acquisition brings its own add-backs, and each one needs support. A reviewer who sees storm-response costs added back will ask whether the storm revenue they produced has been normalized too. The costliest error is a recurring cost presented as one-time. In one illustrative case, $600,000 of recurring costs presented as adjustments puts $4.8 million of implied enterprise value at issue at 8.0×. The arithmetic is simple; the question is whether those costs qualify as adjustments at all. The adjusted EBITDA case Which add-backs hold up

10Are run-rate synergies included anywhere, labeled as forecasts, and within any cap and realization window the credit agreement sets?

Synergies are forecasts. A lender will check whether they are labeled that way and whether they stay within the cap and the time limit the credit agreement allows. A synergy carried past its window, or folded into a brand’s figure without a label, is the add-back most likely to be removed.

11Is all corporate overhead included in adjusted EBITDA, and when the board pack compares brands, does it say whether figures are before or after allocations?

Buyers start with corporate overhead: is all of it in the figure, or has some been left at the holding company? Sponsor fees may be added back; holding-company costs a standalone platform would carry may not. How allocation changes brand comparisons is covered in EBITDA comparability across portfolio companies.

12For each figure, can you show the source records, the definition applied, and who approved any reconciling decision?

This is the question behind the other eleven. A consolidation workbook shows the arithmetic. It rarely shows which definition each brand’s figure was built on, or who decided how a difference between brands was resolved. That record is what turns a reviewer’s question into a quick answer.