Stockouts: What Causes Them and How to Anticipate Them from Your System

On the Friday of Buen Fin, Mexico's biggest shopping weekend, a supermarket chain sends your distribution company a replenishment order for 1,200 stand mixers, the product on the cover of its flyer. The system shows 900 units available, so sales confirms the order and plans to cover the rest with the purchase arriving on Tuesday. On the dock, the pickers find just over 300: some were in a return that was never put away, others are in the wrong location and a full lot is still in another branch's warehouse. Tuesday's purchase does not arrive complete either, because the supplier reported the delay to a buyer who was on vacation. The chain receives an incomplete shipment on the biggest sales weekend of the year, applies the service level penalty and assigns the next replenishment to another supplier.

A stockout happens when a product runs out while demand is still active. It is caused by late orders, unreliable suppliers, demand spikes and stock records that do not match the rack; you anticipate it by watching coverage every day.

What a stockout costs beyond the lost sale

The lost sale is the most visible cost of a stockout, and the others usually weigh more. An out-of-stock product drags costs into the warehouse, transportation, the customer relationship and the data your company uses to plan its next purchases.

Incomplete orders and service level penalties

Retail chains and industrial customers measure their suppliers by the percentage of orders they receive complete and on time. An incomplete shipment usually brings a penalty, a discount on the invoice or a lower score in the supplier evaluation. In peak season, that score decides who gets the replenishment orders.

Rush freight and split shipments

When the product arrives late, the order ships in two or three parts. Each extra shipment takes hours of picking, packing, paperwork and transportation, and it often travels by rush freight or in a half-empty truck. The order's margin shrinks with every extra shipment.

Demand that disappears from the history

While a product is out of stock, the system records zero sales even though customers keep ordering. If your purchasing team calculates the next order with that history, it underestimates demand for the product that runs out most often, and the next stockout is built into that calculation.

The five causes of a stockout

A stockout usually combines several causes. The most frequent are five, and each one leaves a trace in your system's data before the product runs out.

The purchase order that goes out late

When the moment to order depends on what the warehouse clerk sees, the order goes out when stock no longer covers the supplier's lead time. The reorder point settles that decision: it sets, for each product, the stock level at which to order, based on consumption during the lead time and safety stock.

The supplier that delivers late or incomplete

A supplier that promises ten days and delivers in fifteen uses up the safety stock every cycle. Partial deliveries have the same effect with less warning, because the order shows as received and the system treats as covered a product that arrived halfway. Measuring each supplier's on-time rate and fill rate, order by order, shows which ones need a bigger buffer or closer follow-up.

A demand spike nobody anticipated

Promotions, Buen Fin or a large customer that moves its order forward push demand above any average. Safety stock covers normal variation in sales, and a spike planned by the sales team requires an additional scheduled purchase. When sales agrees to a promotion without telling purchasing, the stockout becomes a matter of days.

Phantom inventory: parts the system records and the rack no longer has

Phantom inventory is the parts the system records and the warehouse no longer has: unrecorded issues, undocumented shrinkage, data entry errors at receiving or discrepancies carried over from the last physical count. With inflated stock, the reorder point is never crossed and the purchase order is never issued. The stockout appears when the picker reaches the empty location.

Stock the warehouse has and sales cannot sell

Another part of the inventory exists and is still out of reach for sales: parts reserved for another order, lots in quarantine, damaged goods, product close to expiration or inventory stored in another warehouse. If the system reports everything as a single figure, sales promises parts it cannot deliver. Separating total stock from available stock prevents that promise.

The signals that warn of a stockout

The five causes leave a trail in the data days before the product runs out. These are the signals worth reviewing every day.

Days of coverage shorter than the lead time

Days of coverage divide available stock by average daily demand and tell you how long inventory will last at the current pace. If a product has five days of coverage and its supplier takes eight, an order issued today arrives three days after the product runs out. This is the most direct signal, and in peak season it should be calculated with the demand of the last few weeks.

Products below their reorder point with no open order

A product that crossed its reorder point and has no open purchase order signals a stockout in the making. This list, reviewed every morning by purchasing, turns the signal into an order the same day. If a product stays on it for several days in a row, the cause is usually a held budget, a supplier without credit or a pending approval.

Overdue or partially received purchase orders

An order whose delivery date has passed without receipt, or that arrived with fewer parts than ordered, leaves the product without the coverage the system takes for granted. Reviewing overdue orders and partial deliveries every day lets you call the supplier while there is still time to find an alternative.

Recurring discrepancies in cycle counts

A product that shows discrepancies count after count has a recording problem: a shared location, a wrongly assigned code or an issue that nobody records. Those discrepancies signal phantom inventory, and fixing their cause, in addition to the adjustment, keeps the product from running out while the system still shows stock.

Committed orders that exceed available stock

When confirmed orders add up to more than available stock plus orders in transit, the stockout has already happened on paper, even if the rack still has parts. Detecting it when the order is confirmed lets you negotiate dates with the customer instead of telling them on the day of shipment.

How to anticipate a stockout from your system

Signals are useful when they come from reliable data and someone acts on them. These are the conditions your system and your team must meet.

One stock figure for sales, purchasing and the warehouse

If sales checks one report, purchasing a spreadsheet and the warehouse its own records, each area sees a different inventory. When all three work from the same data, promises based on stock that already left and purchases based on stock that never arrived disappear.

Every transaction recorded when it happens

Real-time stock depends on receiving, put-away, picking, returns and transfers being recorded as they happen. Recording with radio frequency scanners at the dock and at the rack reduces data entry errors and closes the door on phantom inventory.

A risk report with an owner

The signals above fit into a daily report: products with coverage shorter than their lead time, products below their reorder point with no open order, overdue orders and orders that exceed available stock. Each line needs an owner and an action with a date. A report nobody reviews arrives as late as the picker standing in front of the empty location.

Every stockout measured with its cause

Recording every unfilled order with its cause (late order, supplier, demand spike, phantom inventory or unavailable product) shows within a few weeks which one weighs most in your operation. With that information, the adjustment is made where it belongs: in the reorder point, in supplier follow-up, in coordination with sales or in warehouse control.

Reliable stock on an integrated platform

At Oasys we have spent more than 25 years developing management systems for companies in Mexico. Our platform integrates ERP, WMS and TMS, with real-time stock in every company and every warehouse. Sales, purchasing and the warehouse see the same inventory, and purchase orders, receiving and picking are recorded in the same system.

The WMS records receiving and locations with radio frequency, keeps traceability by lot, serial number and expiration date, and controls both owned and consigned inventory. Physical, cycle and full inventories are recorded on the same platform, and reports can be viewed on screen, printed or exported to Excel. Everything runs on our own servers, in a data center.

Frequently asked questions

Which metric should I use to measure stockouts?

The most useful is service level: the percentage of lines or units you shipped complete and on time compared with what your customers ordered. Complement it with the days each product spent at zero. The first measures the effect on the customer and the second points to the products that run out most often.

Does raising safety stock prevent stockouts?

It reduces the ones that come from variation in demand and in the supplier. Late orders, phantom inventory and planned spikes without warning keep causing stockouts even when the buffer grows, and a bigger buffer ties up capital. It is better to fix the cause that weighs most first and adjust the buffer afterward.

How often should stockout signals be reviewed?

Every day for the best-selling products and for those in a promotion, and every week for the rest. In peak season, such as Buen Fin and December, it pays to review all active products every day, because coverage gets used up faster than in a normal month.

If the stockouts in your warehouse are discovered when a picker reaches an empty location, the signals that warned of them were already in your data days before. At Oasys we integrate ERP, WMS and TMS on a single platform, on our own servers, so sales, purchasing and the warehouse work with the same real-time stock and every transaction is recorded when it happens. Find out how.