Ask a warehouse manager what a stockout costs and you get a fast answer: the margin on the units you couldn’t sell. That number is real, and it is also a small fraction of the true cost. The rest of it is not hidden on purpose. It just lands in three other budgets that never get added back to the stockout that caused them, so nobody ever sees the total. Here is where the rest of it goes, and one check you can run on your own data to find out how much of it applies to you.
What does a “cost of stockout” number usually leave out?
Most cost-of-stockout math stops at the register: units you couldn’t sell, times your margin per unit. That is a real number, but it assumes the customer simply waited for you to restock, or bought something else from you instead. Neither is a safe assumption, and the ways it breaks each land in a different department’s numbers.
A customer who cannot get what they ordered does one of a few things: they wait, they buy a substitute from you, they buy the same item from someone else, or they buy it from someone else and quietly start doing more of their shopping there. Only the first two show up anywhere near your stockout report. The other two show up, if they show up at all, as a slightly softer number in next quarter’s sales for that account, long after anyone would think to connect it back to a specific empty shelf. Harvard Business Review put this plainly in a piece on retail out-of-stocks: shoppers do not reliably substitute when the item they want is missing, some of them simply leave, a possibility most stockout arithmetic never enters into the equation at all (Corsten & Gruen, “Stock-Outs Cause Walkouts,” Harvard Business Review, May 2004).
Where does the rest of the cost actually go?
Three places, usually, and none of them are labeled “stockout.”
Expediting. Someone pays for air freight instead of sea freight, or a rush production run, to close the gap faster than the normal cycle would. That cost sits in logistics or in a purchase order variance, filed under “freight” or “urgent buy,” not under the stockout that made it necessary.
Firefighting. A salesperson spends an afternoon calming down an account instead of selling to a new one. A planner reshuffles three other orders to free up a partial shipment. None of that shows up as a cost anywhere. It shows up as time your best people did not spend on something else.
The account that quietly shrinks. This is the one that matters most and gets counted least. An account that had a bad stockout experience does not usually announce it. Their next order is a little smaller, or a little later, or their share of a category that used to be yours moves a few points toward a competitor over the following year. By the time it is visible in the numbers, it looks like ordinary demand softness, not the consequence of a specific event eighteen months back.
Why does this mistake survive for years without getting caught?
Because catching it requires connecting two events that live in different systems and different months. The stockout is logged, if it is logged at all, in inventory or fulfillment. The consequence shows up, much later, in a sales or account-management number that nobody thinks to trace back that far. Each department can see its own piece and conclude it is small. Nobody adds the pieces up, because nobody owns all three budgets at once, so the total the business is actually paying stays invisible even though every part of it is sitting in a system somewhere.
What can you check on your own numbers this week?
You need an export from your ERP and one from your order history. No model, no vendor.
- Pull your worst stockouts from the last year, ranked by how many days the item was unavailable or how many order lines it touched.
- For each one, look up the accounts that tried to order during that window. Compare their order volume in the three months before the stockout to the three months after. A flat or growing account is probably fine. A quiet drop is the cost nobody put in the stockout report.
- Ask logistics for any expedited freight or rush orders in the same week as your worst stockouts. That invoice is real money the stockout caused, sitting in a budget that never gets blamed for it.
- Put all three next to the register number: lost margin, expedite cost, and the account drop-off. You do not need it to be precise. You need it once, to see whether the number everyone quotes is close to the real one or a fraction of it.
None of this requires a model. It requires pulling two exports and looking at them side by side, and it will tell you whether your stockout number is the real cost or the visible third of it.
If the pattern above shows up on your top items, the fix usually is not a bigger safety-stock number across the board. It is knowing which items are actually at risk before they run out. Our per-unit inventory work starts by measuring exactly that on your own purchasing data, before any commitment to build anything.