Books I have co-authored

UK supermarkets still have a lot to learn about Big Data and processes

Looking for out of stock products

Looking for out of stock products


They have yet to solve the replenishment
and restocking problem

For years, retailers have been investing massively in Big Data and analytics. Most of the innovation activity has been on the marketing, insight, and channel side. The industry’s main focus has been on attention-grabbing projects. For example: mobile in-store offers, omnichannel, and drone delivery. The advanced statistical analytics have primarily been used to develop sophisticated consumer profiles rather than optimising processes.

But it seems that the basic challenge of replenishment and maintaining inventory stock levels has not been given the attention it deserves. Market data from US supermarkets shows that, on average, 8% of their products are out of stock. This is consistent with my own experience here in the UK. A survey from last year showed that 21% of UK consumers often experienced in-store stock-outs. 41% of these disappointed shoppers blamed the retailer. And 57% of customers who were unable to purchase products due to stock-outs opted not to buy a substitute, resulting in lost sales.

The stock optimisation problem is far worse during major holidays. On Easter Monday, the meat section at my local Waitrose was brimming with marked down, unsold meat close to its expiration date. During Christmas, most of the major supermarkets had the opposite problem. On 23rd December, a number of shelves were nearly bare in Waitrose as well as Tesco’s largest stores here in the South East of England. In one store, the shelves in the vegetable section were half empty. Basic items such as iceberg lettuce, ginger, broccoli, cucumbers, coriander, etc. were out of stock. Online grocery retailers such as Ocado were also unable to meet demand. Consumers angrily complained about unwanted product substitutes, mistakes, delays, and early pre-orders that suddenly couldn’t be filled. The fact that consumers shop differently for the holidays had obviously eluded the demand planners. For example, people who don’t usually cook themselves will often do so for the holidays.

Stock-outs are a constant annoyance for consumers. I have noticed that one main supermarket near me has a weekly cycle. Out of stock items peak on Mondays. Inventory begins to drop on Sunday and it is not until late Tuesday afternoon that they have managed to get their weekly delivery back on the shelves. Sale items are even worse. During campaigns, this supermarket can be out of stock almost half of the sale period. If I, as a customer, can see these patterns why can’t the retailer identify them as well?

For all their investments in analytics, it seems that supermarkets are drowning in data but are short on insight. A recent survey showed that almost half of UK retail grocery directors use manual replenishment based on gut feeling, while the rest use algorithmic restocking with the option of manual overrides.

Retailers underestimate the costs of out of stock items. If a customer rejects the available substitutes, the retailer loses a sale. And consumers are becoming increasingly selective. Many would not even consider the available “substitutes”. A supermarket that repeatedly fails to stock the consumer’s three favourite products, will eventually lose this customer permanently. The importance of availability is nearly on par with price for most consumers.

The replenishment issue ought to be given higher priority by the Big Data teams. Instead, their focus is often on advanced statistical discovery of hidden buying patterns, price elasticity, and brand preferences. Optimising replenishment is an easier and more straightforward goal. For example: If an item with a weekly replenishment cycle goes out of stock three to four days after delivery, it should be restocked in larger quantities, or be restocked twice per week.

The challenge for the retailers is of course to find the right balance between availability and waste. Unsold fresh food that must be discarded after the expiration date is even more costly than missed sales. Every year, 4 million tons of food is discarded by UK supermarkets. Traditionally, retailers manually mark down items close to the expiration date, which is both time-consuming and expensive. This is an area with significant room for improvement and innovation. The ideal solution would probably be dynamic price tags on each item that could be altered remotely with RFID. The technologies for this (electronic ink etc.) are way too expensive today and will not be competitive in the low margin grocery market for several years. But it would be possible to integrate the expiration date into the product bar codes (or use QR codes). This would provide retailers with a better overview of the remaining shelf life of all products in stock. Products close to the expiration date could be easily marked down without the need to manually mark each item. Electronic shelf labels could display the offers. Combining real-time data of all products close to their expiration date with loyalty programs could be used for generating microtargeted offers.

Selling products close to their expiration date would be a very profitable driver for microtargeting. Customers inside the store or nearby could receive coupons on their smartphones which are valid for a few hours or a day. This would make bargain-hunting customers happy, and substantially reduce the waste generated by discarded products. If the waste from overstocking is reduced, it will be less costly for retailers to keep stock levels high. This will also result in more satisfied customers as they will no longer be frustrated by empty shelves. The potential for both improved ROI and customer satisfaction is substantial.

Comments are closed.