Replacement cost vs average cost: which one do you price against?

A hardware retailer’s margin report said 32%. The shelf price sat below what restocking the item now costs. Both numbers were right — that is the problem.

Every week the vendor cost file lands: thousands of lines, every distributor in its own format, half the costs moved since the week before. Somewhere in that file is the item whose cost went up 155%. The margin report will not show it. The report compares your shelf price against average cost — what you paid for the stock on hand, averaged over every receipt — and average cost moves slowly. Replacement cost is what the next case of the same item costs today. When the two drift apart, the report reads fine and the shelf loses money.

The report reads fine. The shelf is underwater.

Underwater is specific: the shelf price sits below replacement cost while still showing margin against average cost. You sell the unit, book the margin, and then pay more than you collected to put the unit back. The board hears a rate. The P&L loses dollars — every sale of that SKU restocks at a loss, and the faster the item turns, the faster the leak. This is the arithmetic behind "the margin report looks fine and the bank balance disagrees."

Which cost should you price against?

Price against replacement cost, and report against both. Average cost tells you what you paid; replacement cost tells you what staying in stock costs next, and the margin that matters is the one that survives restocking. A simple rule set covers most of it: compute both costs per SKU every week from the vendor cost file; flag any item whose shelf price sits below replacement cost, whatever the reported margin; flag items whose replacement cost moved more than a set threshold, up or down, since the last price change. Falling costs deserve the same attention — pricing to replacement while competitors still hold expensive stock means you are overpriced, and a rule catches that too. The point is not picking one basis forever. It is seeing the gap between the two, every week, per item, before the gap eats the margin.

The gap hides in two places. The first is stale stock: tens of thousands of SKUs priced against costs that no longer exist, because weekly repricing covers only the urgent lines — we have counted 20,000 stale SKUs in a single 25-store file. The second is unit of measure. Buy by the box, sell by the each, formatted differently by every vendor — a 54-cent unit cost becomes an $84.99 shelf price, or quietly erases the margin entirely. Neither shows up in an average-cost margin report. Both show up the week someone reads every line.

Reading every line, every week

By hand, that read does not happen. Checking a full cost file against the price file is days of work, so it becomes sampling, and sampling becomes luck — the 155% increase reaches the shelf only if someone happens to notice. Encoded, it is a batch. The rules your best merchant carries — margin floors by category, replacement-cost checks, the unit-of-measure traps per vendor — become code that reads every line, every week, and surfaces only the violations. This is what we build. We are an AI implementation firm, not a software vendor: there is no platform license and no new system for your team to learn. The rule set is written down, and it is yours.

Average cost is history. Replacement cost is the bill for staying in stock.

What an engagement builds, in what order, at what payback, is on the retail pricing analytics page (zaigo.ai/workflows/retail-pricing-analytics). The retail-specific picture — co-op price files, tag labor, the big box down the road — is on the hardware retail page (zaigo.ai/industries/hardware-retail).

If your margin report and your bank balance tell different stories, the next step is a 30-minute working call: zaigo.ai/book-a-call. Bring last week’s cost file — we will tell you on the call what reading it would take.