Out-of-stocks and overstocks: the two faces of the same stock problem

Two symptoms, one illness

The IHL Group “Inventory Distortion Study” 2026 estimates the worldwide cost of inventory distortion at 1.7 trillion dollars a year, or 6.2% of retail sales. Of that total, 691 billion comes from out-of-stocks (lost sales) and the rest from overstocks (markdowns, waste, tied-up capital). In other words, overstock costs more than out-of-stock at the scale of the sector.

These two symptoms share the same causes, well identified by Corsten and Gruen for out-of-stocks and by DeHoratius and Raman for inventory inaccuracy:

• wrong book stock: too high, it creates out-of-stocks (the order does not go out); too low, it creates overstocks (the order goes out for nothing);
• rough forecasts: the yearly average is too low at peak and too high in the trough;
• ordering parameters never reviewed: reorder point and maximum stock set when the SKU was created;
• a badly kept back room: the product is there but cannot be found, it is reordered, it arrives twice.

Why “ordering more” does not work

The reflex in front of an out-of-stock is to raise the maximum stock. On a low-margin, long-life product, the cost is moderate. On a fresh product, overstock becomes waste within three days (fresh aisle: reduce out-of-stocks without increasing waste). On a seasonal product, it becomes a markdown at the end of the season. On all products, it occupies the back room and the shelf at the expense of other SKUs.

Above all, it does not fix the cause: overstated book stock will keep blocking automatic ordering, whatever the maximum stock.

Track the two indicators together

Situation	Out-of-stock rate	Days of stock	Diagnosis
Well-set SKU	low	in line with delivery lead time	nothing to do
Under-ordering	high	low	raise the reorder point
Overstock	low	high	lower the maximum stock, order more often
Phantom stock	high	high (book)	cycle counting, correct stock
Disorganised back room	high	high (real)	tidy the back room, put products out

The fourth and fifth lines are the ones a single indicator does not show: out-of-stock and overstock at the same time, on the same SKU. They are frequent, and they are fixed neither by ordering more nor by ordering less.

Act on the causes

1. Make book stock reliable on the SKUs that count, through cycle counting (cycle counting: twenty minutes a day).
2. Review ordering parameters of the fifty best sellers once a month, with real sales (automatic ordering and out-of-stocks).
3. Order more often, less at a time when logistics allow: average stock falls and availability rises.
4. Tidy the back room so the delivered product reaches the shelf before being reordered (organising the back room).
5. Add days of stock to the out-of-stock dashboard, SKU by SKU, to see both faces (out-of-stock dashboard).

A worked example (hypothetical)

Take a fictional grocery aisle of 800 SKUs. To bring down a 7% out-of-stock rate, the department manager raises the maximum stock of every SKU by 30%. Three months later, the out-of-stock rate is at 6%, days of stock have risen by a third, the back room is overflowing, and the remaining out-of-stocks are almost all phantom stocks the maximum stock did not touch. By returning to the initial maximum stocks and setting up cycle counting on 50 SKUs, the expected scenario is an out-of-stock rate lower than the one obtained through overstock, with less stock. The figures are illustrative.

What shopper reporting changes

An out-of-stock reported by shoppers with ShelfAlert on a SKU whose book stock is high is the exact signal of phantom stock: the double penalty, out-of-stock and book overstock. That list, produced every day, says where cycle counting should start. 14-day trial, no payment card: see the plan for store managers.

Sources

• IHL Group, “Inventory Distortion Study”, 2026: study overview.
• Corsten, Gruen, “On Shelf Availability”, 2004: summary.
• DeHoratius, Raman, “Inventory Record Inaccuracy: An Empirical Analysis”, Management Science, 2008: abstract.