A seller-commissioned quality-of-earnings report is a streamlining tool because it does the buyer's hardest financial work in advance and puts a defensible number on the table before the negotiation starts. It's also the single best defense against a retrade, because it's an independent analysis the buyer can't easily re-litigate. But its value depends on understanding what a QoE actually is — and it isn't an audit, which is the confusion that trips up first-time sellers and buyers alike.
§ 01 · QoE is not an auditA different question.
| Analysis | The question it answers |
|---|---|
| Audit | Do the statements fairly represent results under GAAP? (backward, technical) |
| Quality of earnings | What is sustainable normalized earnings power, and what would a buyer adjust? (forward, economic) |
The distinction is foundational. An audit is backward-looking and technical — it asks whether the financial statements fairly represent results under GAAP. A quality-of-earnings report is forward-looking and economic — it asks what the agency's sustainable normalized earnings power is, and what adjustments a buyer would rationally make to reach it. For a deal, the QoE question is the one that matters, because a buyer is pricing the future earnings they'll own, not certifying the past. A QoE for an agency deal runs 40–80 pages and works three core areas: revenue quality (is the top line durable, or is it a premium-cycle artifact?), expense normalization (what's the real cost structure under new ownership?), and working-capital analysis. The forensic, buyer-side version of this same analysis is in quality of earnings; here it's the seller-commissioned tool that speeds the deal.
§ 02 · Revenue and contingency normalizationThe core adjustments.
Two normalizations carry the QoE. Contingency bonuses run 6%–10% of direct-bill commissions in a benign-loss year and drop to zero in a bad one, so the conservative default is a three-year trailing average (stripping them entirely is the most conservative). And hard-market top-line distortion — 15%–20% year-over-year growth with no underlying new-business activity — is a pure premium-cycle artifact, so the QoE decomposes revenue into retention, rate premium, new business, and lost business.
The revenue and contingency normalizations are where a QoE separates real earnings power from cyclical noise. Contingency income is volatile — strong in a benign-loss year, zero in a bad one — so capitalizing the most recent year overstates sustainable earnings; the conservative default is a three-year trailing average, and the most conservative treatment strips contingencies out entirely. Revenue gets decomposed rather than taken as a single growing number, because a hard market can produce 15%–20% top-line growth with no new business at all — pure rate inflation that will reverse — so the QoE separates retention, rate premium, new business, and lost business to show how much of the "growth" is durable. A 36-month lost-large-account analysis reconciles to the reported retention rate (high overall retention with several large lost accounts means the retention number is flattering the business). These normalizations are what turn a seller's optimistic number into a defensible one.
§ 03 · Concentration and add-backsWhat the QoE flags and tests.
A QoE also surfaces concentration and disciplines the add-backs. Any single client above ~10% or any top-5 carrier above ~10% gets flagged; a 20% single-client concentration isn't a deal-killer but produces a price discount; and a 50% single-carrier concentration approaches change-of-control review, which is material risk that belongs in the QoE rather than buried in legal diligence. On add-backs, the QoE applies a test to each of five common categories — excess owner comp (legitimate only if the buyer replaces the seller at a market rate), personal auto and travel (legitimate only if genuinely personal), above-market related-party rent (legitimate only if the buyer can exit or renegotiate the lease), non-recurring legal or consulting (legitimate only if the exposure is genuinely closed), and one-time tech investments (legitimate only if the migration is genuinely one-time and complete). The documentation standard is amount, rationale, and supporting documentation for each — without all three, an add-back is an assertion, not a substantiated adjustment, and a buyer's QoE will reverse it.
§ 04 · Four red flagsAnd why the seller commissions it.
Four red flags mark a weak or self-serving QoE, and a buyer reads for all of them: management add-backs accepted without independent validation, revenue treated as a single number rather than decomposed, contingencies baselined at the most-recent year instead of a trailing average, and adjusted EBITDA exceeding reported by more than 25% without commensurate documented adjustments. The reason a seller commissions the QoE themselves — rather than waiting for the buyer's — is the streamlining payoff: a credible, independent QoE on the table at the start collapses the financial diligence (the buyer is confirming an analysis rather than building one) and defends the price against a manufactured retrade, because the buyer can't easily re-litigate add-backs an independent third party already validated and documented. The seller-commissioned QoE is one of the highest-ROI moves in a streamlined sale — it costs real money, but it speeds the close and holds the number, which is exactly the combination the streamlining discipline exists to produce. How it anchors the retrade defense is in retrade defense.
◆
Terminology on this shelf
- Quality of earnings
- A forward-looking economic analysis of sustainable normalized earnings power — not an audit.
- QoE vs. audit
- Audit asks "fair under GAAP?" (backward); QoE asks "sustainable earnings power?" (forward).
- Contingency normalization
- 6%–10% in a good year, zero in a bad one — defaulted to a three-year trailing average.
- Hard-market distortion
- 15%–20% growth with no new business — decomposed into retention, rate, new, and lost.
- Add-back test
- Amount, rationale, and documentation — without all three, it's an assertion, not an adjustment.
- Four red flags
- Unvalidated add-backs, single-number revenue, most-recent contingency baseline, adjusted >25% over reported.