Margin Architecture
Synthetic worked example — not client data
Worked Example · Wholesale DistributionThe biggest customers were the best customers — until the costs found them.
A $54M specialty food and beverage distributor ranks its customer book by revenue and prices deals off invoice margin — price minus product cost. Freight, handling, returns, and $890k of rebates tracked in spreadsheets all sit below the customer line, attributed to nobody. This example models the top-20 book: $23.4M of revenue.
Illustrative scope: the top-20 customer book, $23.4M of revenue inside a $54M distributor. Illustrative fees comprise a $9k Financial Signal Diagnostic and a $28k Financial Signal Fix. Illustrative Diagnostic fees reflect the decision scope, entity and evidence-stream count, evidence readiness, and reconstruction effort. Actual fixed fees are scoped after discovery and evidence-access review. Modeled opportunity amounts are illustrative scenario assumptions informed by the diagnostic findings. They are not realized or guaranteed results.
Gross margin by customer.
Invoice margin says the book sits in a comfortable band, with the large chains a little lower — “volume pricing, that’s normal.” Switch the view to see what happens when freight, handling, returns, and rebates are attributed to the customers that cause them.
Revenue rank was being treated as value rank. The three national chains plus CornerFresh — 45% of the book’s revenue — contribute negative true margin. The six independents, 13% of revenue, generate close to half of all real margin in the book.
The change, customer by customer.
Ordered by how far invoice margin sat from the truth. Shaded rows fall below the level at which a customer covers its share of operating overhead.
Where the margin went.
Four cost pools sat below the customer line. Nothing was hidden — it was simply never assigned to anyone.
The rebate blind spot. $890k of rebates and promotional allowances lived in spreadsheets and hit the ledger quarterly, in arrears, at company level. One chain’s renewal had been priced the previous year off margins that excluded $410k of that chain’s own allowances.
Size is not profitability.
Revenue on the horizontal, true margin on the vertical. The largest accounts in the book sit at or below zero, while the smallest sit highest. UrbanGrocer’s 26 small-format stores take 2,760 tiny drops a year; CornerFresh averages under $1k per drop.
What the work meant.
The diagnostic took three weeks and cost $9,000. The Financial Signal Fix — a customer and segment cost-to-serve model, a monthly on-ledger rebate accrual by customer, a repriced customer tier framework, and the rebuilt Top Customers report now ranked by true margin and owned by finance — cost $28,000. Total illustrative investment: $37,000.
The corrective actions model to roughly $680,000 in the first year: $560k from minimum drop sizes and delivery fees on the small-drop accounts, a rebate restructure at renewal, a price floor with payment-terms enforcement, and $120k from route and drop consolidation now that dispatch data carries a cost. That is an 18.4× ratio of modeled opportunity to fees.
What leadership can act onSet a minimum drop size and delivery fee for UrbanGrocer and CornerFresh
Both order small and often. The contract structure, not the customer, is what makes them unprofitable. Modeled at $310k.
Restructure ValueFoods rebates at renewal
Volume tiers in place of flat promotional allowances, now visible monthly rather than quarterly in arrears. Modeled at $160k.
Hold a price floor on GrandMart and enforce payment terms
At 60-day terms and zero true margin, the largest account in the book funds itself with the distributor’s working capital. Modeled at $90k.
Run route and drop consolidation
Dispatch data existed all along; it had never been costed against customers. Modeled at $120k.
Protect the independents
Thirteen percent of revenue, close to half the real margin, and the segment nobody was defending.
Withdraw the fourth-chain bid
Priced with a fuel-surcharge waiver against UrbanGrocer’s actual drop profile, the new business modeled at roughly −$220k per year. The bid was withdrawn before signing.
The report was not wrong about the invoice. It was wrong about everything that happened after the invoice.
Modeled avoided downside.
The fuel-surcharge-waiver bid for a fourth national chain, modeled against UrbanGrocer’s actual drop profile, priced the new business at approximately −$220k per year. The bid was withdrawn before signing.
This figure is excluded from the $680k headline and from the 18.4× ratio. It is a counterfactual: it depends on a decision that was never executed, and it describes business the distributor does not have. It is shown here because the finding mattered, not because it can be banked.
Wholesale food distribution gross margins benchmark at roughly 10–30% depending on category and service model, which places this example’s 18.8% invoice margin mid-band — the reason nothing looked wrong at company level. Industry net margins are famously thin, which is why cost-to-serve analysis treats per-drop economics as the decisive lens.
Sources: Unleashed food distribution industry analysis; FullRatio industry margin data; CSIMarket wholesale profitability ratios. Accessed 2026.
Release controls
- Synthetic worked example — not client data. Riverline Specialty Distribution is an invented company.
- Invented company, real method. The analytical method and the arithmetic are real; the company is not.
- Modeled opportunity, not realized or guaranteed results.
- Impact estimates include illustrative scenario assumptions informed by the diagnostic findings.
- Fees shown reflect this illustrative scope only.
- The Financial Signal Diagnostic and the Financial Signal Fix are separate phases.
Start with the report or number leadership no longer fully trusts.
If a margin, profitability, pricing, cost, or revenue-unit signal is important enough to act on but not trusted enough to act on confidently, that is the conversation.
Cost-to-serve inversion appears wherever the cost of serving a customer is not carried on the same line as the revenue from that customer. Distribution makes it easy to see because the costs are physical — trucks, drops, miles. The pattern is not confined to distribution.