An earn-out is the tool for risk you can see but can't yet value. It earns its place when diligence turns up single-client concentration above 25%, retention below 85%, a key-person dependency, producer-ownership concerns, or carrier-appointment consent uncertainty — anything where the book's future behavior, not its past, is what you're really pricing. The structure converts that uncertainty into a contingent payment the seller earns only if the risk doesn't materialize.
§ 01 · The metricClean to dispute-prone.
| Metric | Manipulation resistance | Best use |
|---|---|---|
| Policy retention | Highest — AMS-based, unit count | The default for most agency earn-outs |
| Revenue retention | Lower — gameable in hard markets | When economic reality must be captured |
| Specific-client retention | High — named clients, binary/scaled | Concentration-driven earn-outs |
| EBITDA-based | Lowest — post-close decisions distort it | Rarely; sellers distrust it |
Policy retention (unit count) is the cleanest because it's AMS-based, resistant to manipulation, and tied directly to the underlying risk. Revenue retention captures economics but distorts in a hard market. Specific-client retention — named clients at 12 and 24-month checkpoints — is the right tool for a concentration-driven earn-out. EBITDA is the most dispute-prone, because countless post-close buyer decisions move it and sellers rightly distrust it. A common hybrid splits the earn-out 50/50 between overall policy retention and the retention of specified top clients, capturing systemic and concentration risk in one structure.
§ 02 · Duration and sizingHow much, and when.
Match duration to the risk: a single 12-month measurement for lower-risk deals with modest retention concerns; a 24-month structure split into 12 and 24-month checkpoints for meaningful risk — concentration, weak retention, producer-ownership concerns; 36 months only for genuinely complex, long-cycle commercial books, and rare in small-agency deals. Size it around 70–80% cash at close, 10–20% at the 12-month checkpoint, and 5–15% at 24 months. High-risk deals can drop to 60% cash at close, but below 50% cash is rare and usually signals the deal shouldn't happen in that structure at all.
§ 03 · The five pushbacksAnd their middle grounds.
Sellers raise the same five objections, each with a workable answer. "I won't be in control" — choose the metric that minimizes buyer influence (policy retention over revenue over EBITDA) and add ordinary-course-of-business covenants. "You can depress the metric" — add explicit anti-manipulation clauses prohibiting intentional client-shedding or carrier non-renewal designed to force departures. "I should be paid for the risk I'm carrying" — a 5–10% premium on the earn-out portion versus a comparable all-cash deal. "What about your solvency?" — escrow, a parent guarantee, or a letter of credit for the earn-out amount. "Measurement will be a fight" — a worked-example methodology, an arbitration mechanism, and an independent accountant or industry-specialist adjudicator named at signing.
An earn-out prices performance; a holdback secures indemnification. They're different tools and they layer — and an indemnification claim can be netted against an unpaid earn-out rather than requiring the seller to write a check, a natural cash-flow offset between the two.
§ 04 · Layering the toolsOne risk, one instrument.
The discipline that makes a complex deal hold is matching each tool to one dimension of risk rather than asking any single mechanism to carry everything. A concentrated book with weak producer contracts is the canonical case: an earn-out on concentrated-client retention prices the performance risk, a holdback secures the indemnification, specific reps cover the producer agreements, and a seller transition plan aligns the hand-off — four instruments, four risks, no overlap. On larger deals ($3M–$5M+ purchase price), representations-and-warranties insurance can substitute for some earn-out protection, shifting reps-breach risk from the seller's cash-at-risk to an insurer, though it's less common in sub-$5M small-agency deals. Choose the metric the seller can't game, size the payout so cash-at-close stays healthy, and let the earn-out do exactly one job: turn a risk you can see but can't yet value into a shared bet on the outcome.
◆
Terminology on this shelf
- Earn-out
- A performance-contingent payment based on post-close metrics — the tool for risk you can see but can't price at close.
- Policy-retention metric
- The cleanest earn-out measure — AMS-based unit count, resistant to manipulation.
- Specific-client retention
- Named-client checkpoints at 12 and 24 months — the right metric for a concentration-driven earn-out.
- Anti-manipulation clause
- A covenant barring the buyer from intentionally depressing the earn-out metric.
- Holdback
- Escrowed consideration released on defined triggers (often indemnification) — distinct from, and layered with, an earn-out.
- R&W insurance
- Reps-and-warranties insurance that shifts breach risk to an insurer — used on larger deals.