A file review is where customer due diligence stops being a spreadsheet and becomes a read of how the agency actually serves its clients. The work is sampling discipline plus pattern recognition: pull a representative set, read each file the same way, and let the distribution of what you find drive both price and protection.
§ 01 · The sampleThree layers, 60–100 files.
A small-agency review runs 60 to 100 files in three layers. The top 15–20 by commission get 100% coverage — they carry the concentration risk and the most material valuation impact. A random 25–50, stratified by line, producer, and tenure, surfaces systematic problems the weighted sample misses. And a claims-weighted oversample covers every meaningful claim in the trailing 24 months — roughly 10 files in a small agency, 10–20% of claims in a larger one — because the claim moment is where retention is won or lost. The principal buyer reads certain files personally: every top-20 file (read, not skimmed), every claims file where the claim exceeded $50K (or $500K at larger scale), anything the analyst team flags for E&O, and a random 10-file calibration pull.
§ 02 · Four dimensions per fileWhat to read.
Each file gets audited on four parallel dimensions. Policy documentation — a signed application, a current dec page within the last renewal cycle, endorsements in effect, carrier correspondence, proof of delivery, and signed rejection forms for declined coverages like uninsured motorist or higher deductibles. Claims history. Client communication — a client touched once a year at renewal is a flight risk; one touched four to six times a year with proactive coverage reviews is an asset. E&O compliance. The documentation red flags are familiar: missing signed applications (the classic E&O exposure), no documented declinations, dec pages two or three renewal cycles stale, "we talked about it" with nothing in writing, records scattered rather than consolidated. The claims red flags run parallel: no intake documentation, long communication gaps during active claims, carrier-side-only correspondence with no agency advocacy visible, and repeat-claim clients who lapsed without any retention effort.
§ 03 · Reading the aggregateThree tiers, three responses.
| Pattern | Clean files | Buyer response |
|---|---|---|
| Clean sample | 80%+ | Priceable / footnotable — active comms, no E&O concerns |
| Mixed sample | 50–80% | Fixable post-close — targeted indemnification + negative multiple adjustment |
| Deficient sample | <50% | Disclosure event — reprice, expand reps/indemnification, or walk |
The tiers turn on whether the gaps are isolated or systematic. A mixed sample with problems concentrated on one producer, line, or service team is fixable — you scope a targeted indemnification and adjust the multiple. A deficient sample with gaps spread across the book is a different animal: the deficiency is the deal's defining fact, and the response is to reprice, expand the reps and indemnification materially, or walk.
§ 04 · Sizing the protectionsFrom findings to the agreement.
E&O findings in the file review are the single highest source of post-close indemnification claims. Let them, not a boilerplate template, size the reps, the indemnification basket, and the tail coverage — the file review is the evidence base for the protective half of the purchase agreement.
Four agreement protections get sized off the findings. Specific representations — file completeness, no undisclosed E&O, claims handled to industry standard, coverage recommendations appropriate and documented. An indemnification basket and cap with sub-baskets carved out for pre-close coverage recommendations and documentation gaps. Tail E&O coverage, its limit and duration set by what the review surfaced. And pre-close remediation requirements — missing signed rejections, stale dec pages, AMS reorganization — that transfer the cleanup cost back to the seller. Read the files, score the tiers, and the agreement writes itself from evidence rather than assumption.
◆
Terminology on this shelf
- Three-layer sample
- Top-20 by commission (full coverage) + a stratified random 25–50 + a claims-weighted oversample.
- Four audit dimensions
- Policy documentation, claims history, client communication, and E&O compliance — read in every file.
- Communication cadence
- A retention signal: once-a-year (renewal only) is flight risk; 4–6 proactive touches is an asset.
- Aggregate-pattern tiers
- Clean (80%+), mixed (50–80%), deficient (<50%) — each with a distinct pricing and protection response.
- Tail E&O coverage
- Post-close errors-and-omissions coverage whose limit and duration are set by file-review findings.
- Indemnification basket
- The reps reserve, with sub-baskets sized to documentation gaps and coverage-recommendation exposure.