Stockout

A stockout is the moment an item you need to sell or use has no stock available, so an order, a dispatch or a production run must wait or be refused. In Hindi it is स्टॉक से बाहर (Hinkhoj).

A backorder is different: the customer can still order and is given a delivery date. A low-stock warning is not a stockout either. It is the signal that comes before one.

Related terms:

• Safety stock: the buffer held against surprises. See safety stock calculation.
• Reorder point: the stock level that triggers a new order. See reorder point formula.
• Lead time: the days between placing an order and receiving it.
• SKU (stock keeping unit): one distinct item code.

A stockout can mean zero on hand, zero available, or zero at the pick location while the system still shows stock. Keep them apart, because each has a different cause and a different person who can fix it.

Three kinds of zero behind a stockout

In a 2002 compilation of grocery retail studies, 72% of stockouts began inside the store and 28% upstream of it. Gruen, Corsten and Bharadwaj's report for the Grocery Manufacturers of America gives 47% from store ordering and forecasting, 25% from stock in the store but not on the shelf, and 28% from upstream causes (Exhibit 2 of its executive summary). The two store-side shares sum to 72%.

The report covered 52 studies in 29 countries and fast-moving consumer goods retail, so it is not Indian business-to-business data. The inference is that distributors with many fast-moving SKUs see the same pattern, but a business's own log should confirm it.

Record errors feed this directly. DeHoratius and Raman found 65% of about 370,000 inventory records inaccurate in 37 stores of one retailer (Management Science, 2008). In a study of seven European retailers, about 60% of audited SKUs had record errors, and correcting them lifted sales by 3.8% to 8.4% across the retailers, 6.0% on average (ECR Retail Loss). Fast-moving items gained most from the corrections.

Recommendation: count the fastest-moving items first. See cycle counting.

Measure the rate as SKU-days at zero available (zero on hand, or zero available after commitments, as the system shows) divided by active SKUs times days in the period, multiplied by 100. Cost one stockout as units demanded, times the share of buyers who walk away, times margin per unit. Both are models, not official figures.

Rate, worked example (illustrative): 120 active SKUs over 30 days is 3,600 SKU-days. If 108 SKU-days sat at zero available, the rate is 108 ÷ 3,600 = 3.0%. Always state the base. The 2002 report's 8.3% average comes from shelf audits at a single point in time, so it should not be compared with a SKU-day rate.

Empty-bin events: the third zero is invisible to this rate, because the system still shows stock. Record those separately as SKU-days where a count found the bin empty, and report that figure next to the rate.

Cost, worked example (illustrative): demand of 40 units a day for a 5-day stockout is 200 units. The 2002 report found 31% of shoppers bought elsewhere and 9% did not buy, so 40% is used as lost. That gives 80 units. At ₹150 margin a unit, lost margin is 80 × ₹150 = ₹12,000. Those are retail shoppers. After a quarter of logging, replace 40% with your own share of lost orders.

A stockout log should record the item, the dates, the quantities and one cause code, so you can total causes by quarter. Copy this layout into a spreadsheet; the example row reuses the figures from the cost example above.

Pass/fail tests for each row:

• Days at zero equals End minus Start.
• Cause code is one of A, B, C or D. A row that gives no cause fails.
• Bin empty at count is Yes or No, and a Yes row carries code B.

Four cause codes split the study's store-ordering, in-store stock and upstream groups as follows:

• A ordering or forecast (store ordering and forecasting)
• B record error or stock not at the pick location (in the store, not on the shelf)
• C supplier late (upstream)
• D stock already promised to another order (in this reading, store ordering and forecasting)

We built Arka Inventory so each kind of zero shows up before an order is lost. For zero on hand, our inventory alerts flag items approaching or below minimum stock and tell users when back-ordered stock arrives; as of October 2026, alerts and purchase orders are included in the Basic plan and above (pricing). Set thresholds on sellable stock as described in low stock alert.

For zero available, Commit Inventory lets you commit stock to orders and allocate or reallocate it by delivery date, so sales sees what can be promised, not only what is physically on hand. For zero in the bin, we show stock by warehouse, and by bin on our Enterprise plan, so a counter can compare the record with the shelf. Follow cycle counting to schedule counts.