A stockout is a situation in which demand for a product cannot be met from available inventory at the moment a customer wants to buy it. In inventory management within e-commerce, a stockout means the product shows as unavailable, the order is placed on backorder, or, in the worst case, the store sells units it does not actually have.
Stockouts are rarely caused by a single event. They usually result from the interaction of demand variability, supplier lead time, replenishment rules and inaccurate stock records, which is why preventing them is a planning discipline rather than a matter of ordering more.
Why stockouts happen
Every replenishment cycle contains a period of risk: the time between placing a purchase order and receiving the goods, known as the lead time. If demand during that period exceeds the stock on hand plus any buffer, the item runs out. The main root causes fall into a few recognizable groups.
| Cause | Mechanism | Typical signal |
|---|---|---|
| Forecast error | Demand is higher than planned | Sales spike after a campaign or a viral mention |
| Lead time variability | Supplier delivers later than expected | Repeated late purchase orders |
| Insufficient safety stock | Buffer too small for normal variation | Frequent stockouts of steady sellers |
| Inaccurate inventory records | System shows stock that does not exist | Orders fail at picking despite available status |
| Channel synchronization delays | Several channels sell the same units | Overselling across marketplaces |
The case of phantom inventory deserves particular attention. When records show units that were lost, damaged or misplaced, the replenishment system sees no reason to reorder, so the stockout lasts until someone physically counts the shelf.
Measuring stockouts and service levels
A store cannot manage what it does not measure, and stockouts are surprisingly easy to underestimate because lost sales do not appear in order data. Several complementary metrics capture different aspects of availability.
| Metric | Definition | What it reveals |
|---|---|---|
| Stockout rate | Share of SKUs or days with zero available stock | How often products are unavailable |
| Fill rate | Share of ordered units shipped from stock | How much demand is served immediately |
| Cycle service level | Probability of no stockout during a replenishment cycle | How well safety stock is sized |
| Backorder volume | Units ordered but awaiting stock | Demand the store still hopes to keep |
| Estimated lost sales | Expected sales during out-of-stock periods | Revenue impact of unavailability |
Estimating lost sales requires a model of normal demand, for example the average daily sales before the stockout multiplied by the days without stock. The figure is an approximation, but it turns availability into money, which helps when deciding how much to invest in buffer inventory.
How safety stock and reorder points prevent stockouts
The standard defense is a reorder point with safety stock. The reorder point equals expected demand during the lead time plus a buffer. A common formula for that buffer, assuming normally distributed demand and a fixed lead time, is safety stock = z × σd × √L, where σd is the standard deviation of daily demand, L is the lead time in days and z reflects the target service level.
| Target cycle service level | Safety factor z | Practical meaning |
|---|---|---|
| 90% | About 1.28 | Roughly one stockout in ten cycles |
| 95% | About 1.65 | Common default for regular products |
| 98% | About 2.05 | Important products with loyal buyers |
| 99% | About 2.33 | Critical items where stockouts are very costly |
The relationship is strongly nonlinear: moving from 95% to 99% requires about 40% more safety stock, which ties up capital and warehouse space. For this reason, many stores set different service levels by product class, protecting best sellers heavily and accepting occasional stockouts on slow, low-margin items.
Stockout responses and related inventory concepts
When a stockout does occur, the store must decide how to present the unavailable product. Each option balances revenue preservation against the risk of disappointing customers with long or uncertain waits.
| Response | How it works | Best used when |
|---|---|---|
| Backorder | Accept the order and ship when stock arrives | Restock date is reliable and close |
| Preorder | Take orders before the first delivery | New or seasonal products with a fixed launch |
| Back-in-stock alert | Collect emails and notify on restock | Restock date is uncertain |
| Substitute suggestion | Recommend comparable available products | Close alternatives exist |
| Dropshipping fallback | Supplier ships directly to the customer | Supplier holds stock and accepts single orders |
Stockout risk also shapes the choice of fulfillment model. A 3PL warehouse adds inbound transit time that must be included in the lead time, while dropshipping shifts stock risk to the supplier but removes visibility into how much stock really exists.
Business impact and common mistakes
The cost of a stockout extends well beyond the missed order. Customers who find a product unavailable may switch to a competitor permanently, paid advertising keeps sending traffic to unavailable pages, and marketplaces may lower the ranking of listings that frequently go out of stock.
- Using average lead time only while ignoring its variability underestimates the required buffer.
- Continuing ads for unavailable products wastes budget and frustrates visitors.
- Selling the same units on several channels without real-time synchronization causes overselling and cancellations.
- Skipping regular cycle counts allows phantom inventory to block reordering.
- Applying one service level to every SKU overstocks slow items and understocks best sellers.
A non-technical view
A stockout is simply running out of something people want to buy. Imagine a bakery that sells its last loaf at noon while customers keep arriving until evening. Every later customer leaves empty-handed, and some may not come back.
Online stores face the same problem. They avoid it by keeping a small reserve and reordering early enough. The reserve covers surprises, such as a sudden rush or a late delivery from the supplier.
A concrete example
An online pet supply store sells a popular dog food bag at about 40 units per day, with a standard deviation of 10 units. The supplier delivers in nine days. The store had set its reorder point at 360 units, exactly the expected demand during the lead time, and it ran out several times each quarter.
The planning team recalculated using a 95% service level. Safety stock became 1.65 × 10 × 3, or about 50 units, which raised the reorder point to roughly 410 units. It also added a back-in-stock alert and paused ads automatically whenever available stock fell to zero.
In the following quarter, the product did not run out once, even during a promotional week. The extra 50 units cost far less to hold than the sales and repeat customers previously lost during each stockout.