How to track stock in real time
If grocery retailers had better visibility of stock and near perfect availability accuracy, the sales wins can run into the millions, says Alex Considine Tong, Chief Product Officer, Retail Insight.
Can any retailer truly answer the question, what is on the shelf right now? There is still no certainty, despite greater visibility than ever, from advanced forecasting models that can predict demand at SKU and store level to automated replenishment systems that generate orders precisely; all augmented by supply chains with data from depot to delivery.
Subscribe to TRBThis is not a data problem in the normal sense because retailers are not short of information. What is happening is, much of that information is delayed, fragmented or inaccurate due other factors such as theft, damage or misplaced. What appears in systems as available is often simply an assumption based on past events rather than a reflection of current reality. This gap between inventory recorded on the system and physical shelf availability has become one of the most important sources of lost sales in grocery retail today.
The scale of the problem is increasingly well understood. In the US, 78% of grocery shoppers said they had encountered out of stocks in store. As a result, a quarter of them went elsewhere to find the item. In the UK, the picture is equally stark. One in five grocery trips now includes at least one missing item, equating to hundreds of millions of disrupted shopping missions each year and billions in displaced sales. Almost half of consumers have switched or added a retailer due to availability issues, while the majority now shop across multiple stores to complete a single basket.
The impact on retailers is getting worse, not because shopper behaviour has fundamentally changed, but because their budgets are being squeezed and they are searching harder for value. However, it is availability that is increasingly the deciding factor at the point of purchase. If a product is not on the shelf, the question of value becomes irrelevant.
Every gap between recorded stock and physical availability is a potential missed sale, often with little chance of recovery. And all these gaps add up to significant margin erosion, as demand is either lost entirely or fulfilled elsewhere. Improving stock accuracy, therefore, is not just about protecting loyalty, but about systematically recovering revenue that would otherwise be invisible and unreclaimed.
The cause of the problem is not typically a supply chain failure because in many cases, the challenge is that the signals connecting supply, store and shelf are not synchronised in real time. Inaccurate inventory then compounds. For instance, deliveries may be received late or partially checked in or cases sit in the backroom longer than expected due to labour constraints. Stock is moved manually without being recorded immediately while shrinkage, damage and misplaced items gradually erode accuracy. At the point of sale, inventory positions are updated in batches rather than continuously.
Arguably, taken individually, these are manageable variances but across hundreds or thousands of stores, they compound into a persistent gap between recorded stock and physical reality. And this is where latency becomes critical; even a delay of a few hours can mean the difference between a shelf being replenished in time to capture demand, or remaining empty during peak trading. Without continuous visibility, retailers are effectively managing stores based on a historical view of inventory rather than a live one.
Real-time stock tracking changes this dynamic by introducing a continuous, reconciled view of inventory across the store. Rather than relying on periodic updates, it integrates signals from goods receiving, stock movements and point of sale activity to maintain an up to date representation of what is actually available for sale. And, rather than treat all discrepancies the same, it identifies exceptions, prioritises them based on commercial impact and directs attention to where intervention is needed most.
This has several immediate effects. First, it enables earlier detection of phantom inventory, where systems indicate stock is present but it cannot be found on the shelf. By identifying these gaps in near real time, store teams can intervene before lost sales accumulate.
Second, it allows retailers to distinguish more clearly between supply side and execution side issues. If stock has not arrived, the response sits with the supply chain. If it has arrived but is not on the shelf, the issue is operational within the store. This distinction will affect how the problem is resolved.
Third, it improves labour allocation. Instead of reactive, manual checks, store teams can focus on the specific tasks that will have the greatest commercial impact, reducing wasted effort and improving productivity.
Real time stock visibility is the enabler of a wider range of strategic initiatives. Retail media performance, for example, is increasingly dependent on product availability. Promoting items that are not reliably on the shelf undermines both campaign effectiveness and supplier confidence. Similarly, AI-driven automation relies on accurate, timely data. Models trained on distorted or delayed inventory signals will produce suboptimal outcomes, regardless of their sophistication.
Even the customer experience is directly shaped by this layer. Poor availability leads to substitutions, cancelled orders and fragmented baskets, all of which erode trust. In a market where shoppers are already splitting their spend across multiple retailers, these moments matter.
Inventory accuracy must reflect reality continuously, not periodically and it must inform action, not just analysis. Real-time stock tracking closes the loop between what the system believes and what the customer experiences.
