Cost rose. Price didn’t.
A confirmed supplier step-up was absorbed instead of passed through to one customer.
Send sales + purchase lines. We cross-check every customer × SKU and return only the exceptions worth touching — conservative annual dollars, the next action, and the rows that prove it.
A person can spot an obviously bad sale. What is hard is comparing a year of receipts against every customer’s price history, continuously, across the whole book.
A confirmed supplier step-up was absorbed instead of passed through to one customer.
A special price stayed flat while your cost or comparable customer prices moved around it.
No model. No peer benchmark. The sell price is simply below the matched acquisition cost.
You do not need another dashboard.
You need a Monday-morning list.
The report ranks the relationships worth attention, tells you why each fired, gives a conservative annual figure, and tells you what to check or change next. Then it puts the invoice and receipt rows underneath.
The audit is deliberately harder to exaggerate than a normal “margin opportunity” report.
Accounting-backed losses are one pool. Benchmark- or history-based opportunities are another. They are never added together.
The calculation, invoice rows, receipt rows, dates, costs and prices sit under the recommendation.
Targets are capped and recovery is discounted. The report is designed to survive a controller trying to disprove it.
No ERP access, no write-back, no customer contact. The audit identifies; your team decides.
Sales lines + purchase/receipt lines, ideally 12 months. CSV, Excel or zip. Odd column names are fine.
What mapped, what cannot be trusted, and how much sales volume has usable cost coverage. No dollar findings yet.
If the files support it, we send a payment link. The full self-contained HTML audit arrives within 10 business days.
No. If you can export sales history and purchase/receipt history, that is the requirement. We never log into or write back to your system.
Messy is expected. The free check exists to identify what mapped, what did not, and which parts cannot support trustworthy dollar findings.
Sales: date, customer, SKU, quantity, price. Purchases: date, SKU, quantity, cost. Extra columns are fine. Landed cost is preferable when available.
This is not a replacement for it. The audit looks across histories for customer×SKU relationships where cost movement, frozen pricing or landed cost creates a specific exception worth reviewing.
Then it stays a review finding. The report cannot see contract terms unless they are represented in the files; that is why benchmark/history findings are separated from high-confidence accounting findings.