Stockout: What It Is and How to Prevent It

What is a stockout

A stockout is when a product runs out and no units are available when someone asks for them. In a store it is the empty gap on the shelf; in a warehouse, the order that cannot be shipped; in a factory, the missing raw material that forces the line to stop.

The cost goes beyond the lost sale. The customer buys another brand or goes to another store, rush orders have to be placed and, in production, a stoppage affects everything downstream.

Out of stock in the warehouse vs. on the shelf

It helps to separate two cases. When the warehouse is out of stock, there is no inventory anywhere. When the shelf is out of stock, there is product in the back room or the warehouse, but not on the shelf, because it was not restocked in time. The inventory system says the item is in stock and the customer cannot find it. This second case is the hardest to see from the data.

Common causes of stockouts

• A demand forecast that falls short because of a promotion, the season or the weather.
• Supplier orders placed late, or lead times longer than expected.
• System inventory that does not match what is actually there.
• Shelf restocking that comes too late.

How do you detect a stockout

• Inventory data: stock reaches zero or drops below the set minimum.
• Sales data: a product that sells every day suddenly stops selling, often because it is not on the shelf.
• Store checks: staff walk the aisles and note the gaps.
• Images: cameras that look at the shelf and a computer vision model that detects empty gaps and alerts the team.

How to prevent stockouts

There are two fronts. The preventive one is forecasting demand better and adjusting orders, inventory and production to that forecast. The reactive one is spotting the gap early and restocking before the customer notices.

An example

A supermarket runs a promotion on a soft drink. Sales rise more than expected and by mid-afternoon the shelf is empty, even though there are cases left in the back room. If an alert tells the team when the gap appears, the shelf is restocked in minutes instead of the next day.

Stockouts at DPI

DPI's computer vision for retail solution includes stockout detection among its applications. It works with existing video surveillance cameras, processes images locally on an NVIDIA Jetson edge device and shows alerts in dashboards, to tackle late restocking and the lack of alerts when consumption spikes. In Buffet Inteligente, it detects empty trays and alerts the kitchen team.

In DPI's machine learning consulting, demand forecasting is used to anticipate product sales and adjust inventory and production, to avoid stockouts or extra costs.