You fund a price you cannot see delivered
The manufacturer’s version of the foodservice problem is the mirror of the distributor’s. The distributor worries that deviated cost is the least governed transaction in their business. You have the same worry from the other side, with less information: the price was yours, the sale was theirs, and the only record of it is the claim.
This is not an accusation of bad faith. Most bill-back error is structural — a stale deviation applied after it expired, a location coded to the wrong operator, the same case claimed twice under two agreements that both legitimately exist. It is error that a sampling audit cannot find, because the errors are small, numerous and spread across every distributor file you receive.
Sampling is not a control
Validating five percent of claim lines tells you the error rate. It does not recover the money. The economics only work when validation is line-level and automatic: every claim line matched against a live authorisation, with the operator resolved to your customer master before the match is attempted. Manufacturers who move from sampling to full-coverage validation typically recover three to seven percent of submitted claim value — not because their distributors were cheating, but because nobody had ever checked all of it.
The identity problem is the real blocker
Full-coverage validation fails on identity long before it fails on rules. A chain location arrives on one distributor’s file under their internal account code, on another’s under a different one, and on neither under yours. Until the operator is resolved, the authorisation cannot be found, and the line falls to manual review — which is where full coverage quietly becomes sampling again.
What good looks like
- 100% of claim lines validated at line level, against authorisations that were live on the date of sale.
- Operator identity resolved above 98% — this is the number that decides whether full coverage is affordable.
- Gross-to-net inside five days of period end, computed from validated lines rather than an accrual estimate that a later true-up will contradict.
How RevUpra runs this
Deviation authorisations are held as live, dated entitlements. Distributor claim files are ingested in their own formats and normalised, with operator identifiers cross-referenced to your customer master so the match rate is high enough for full coverage to be economic. Every line is validated against the authorisation that was in force on the sale date, overlapping claims across deviations, group agreements and allowances are detected before settlement, and what remains is settled or disputed with the evidence attached. Gross-to-net accrues from validated lines, so the number you report at period end is the number that settles.