M&A diligence
Buying a marketplace? Audit the refund path before the deal closes
Diligence on marketplaces scrutinises unit economics, cohorts and churn with forensic depth — then accepts payments health as "Stripe handles it". Meanwhile the target may be funding every refund twice, absorbing dispute losses invisibly, and carrying six months of stranded transfers in churned-seller accounts nobody counted as liabilities. A pre-close Connect audit reframes silent leakage as a quantifiable adjustment item — and occasionally finds nothing, which is also worth knowing before signing.
The diligence checklist
Four probes, ordered by information yield: enumerate every refund path in the target’s repositories (grep refunds.create; count call sites missing either flag); run a consented 90-day lookback scan quantifying current leakage; pull dispute-loss history checking recovery coverage; review FX settlement configuration across currencies. Each probe produces numbers, not narratives — the currency diligence actually trades in.
How findings move price
Modelled framing: measured annualised leakage times your multiple becomes an EV adjustment conversation, not an argument. A $60k/yr leak at a 4× revenue multiple argues for a materially different number than the target’s clean P&L implied. Even unresolved-before-close findings convert into escrow holdbacks or post-closing covenants — mechanisms exist precisely because diligence finds things.
Reps, warranties, and honesty about limits
Direction only, explicitly not legal advice: payment-integrity reps benefit both sides — sellers with clean scans gain selling points; buyers obtain recourse for discovered gaps. Counsel drafts language; the audit supplies factual basis. Scans run only with explicit written consent from authorised officers; surprise audits poison deals.
Day-one post-close
Connect monitoring before integration teams touch anything: integrations routinely break payment assumptions, and baseline-first lets you attribute post-close anomalies to pre-existing leakage versus migration damage. The first board update after close should contain a leakage section because the data already exists — not because someone spent a month assembling it.
What FeeGuard contributes
Read-only scans with consent, evidence-grade findings, historical lookbacks covering the periods reps will reference, and monitoring that continues into ownership seamlessly. Diligence-speed matters: scanner-first results arrive in hours; key-connected depth follows within days.
Common questions
Will the target cooperate?
Framed as operational due diligence with written consent, routinely. Framed as ambush, rightly not. Process design is part of the work.
How fast is a scan?
Minutes once credentials exist; the negotiation is the slow part. Paste-based scanner results arrive even sooner for preliminary sizing.