Finding dead stock, and deciding what to do about it
Dead stock is money on a shelf. The hard part is not spotting it — it is deciding what it is now worth.
2 min read
Dead stock is inventory that is not selling and is not going to. It is the most expensive thing in a stockroom because its cost is invisible: it does not appear on the profit and loss, it appears as an asset, and it goes on appearing as an asset for as long as you leave it there.
The metrics that find it
| Metric | What it tells you |
|---|---|
| Days since last sale | The bluntest and most useful single signal |
| Sell-through | What proportion of what you had has actually gone |
| Inventory turnover | How many times the stock cycled in the period |
| GMROI | Gross profit returned per unit of money tied up |
| Stock aging | How long the units on hand have been on hand |
Two rules for computing them honestly
Divide by average inventory, not by closing stock
Turnover and GMROI both divide by inventory value. Using today’s value makes both meaningless for anything seasonal: buy in October, sell through December, and a January reading divides a quarter of cost of goods by an empty shelf, producing a spectacular turnover figure for a product that is currently not there. Use the average of the value at the start of the window and the value now.
Blank is not zero
Where nothing was held, these ratios have no value — not a value of zero. Zero reads as "turns over badly", which will put a product you have never stocked at the top of your problem list. Show it blank and sort it last.
Rank by profit, not revenue
ABC classification conventionally ranks by revenue, but that convention exists because most systems do not know their cost of goods per sale. If yours does, rank by gross profit. The difference is not academic: a high-revenue, low-margin line and a low-revenue, high-margin line are opposite decisions, and revenue ranking recommends protecting the wrong one.
Then the harder question
Identifying dead stock is arithmetic. Deciding what it is worth now is judgement, and the two should not be automated together.
A useful discipline is to floor any automatically proposed markdown at cost. Clearing below cost is often the right decision — the cash and the shelf space are worth more than the loss — but it is a decision someone should take deliberately. A queue that proposes selling at a loss on its own gets approved by habit within a month.