We all know what it’s like. You go into the supermarket or the pharmacy or your favourite fashion store and there it is, the gap on the shelf that spells stock-out. We have all then followed the usual routine of button-holing a sales assistant and asking if he/she could check in the back room and then we wait for the inevitable return and those dread words: ‘Sorry, it’s out of stock.’
My personal favourite was the young man working in Tesco who after three visits on three consecutive days told me with admirable honesty that: ‘The delivery truck has gone AWOL and we don’t know what will be coming in – or when.’
Hardly surprising then that a recent clothing sector survey for IT specialist Novasoft found that availability can often be as low as 38 per cent with typically a quarter of size options missing from the shelves. In the grocery sector stockouts are often quoted as averaging around seven to nine per cent although double figures are commonplace. Measuring stockouts has, until recently, been a fairly ad hoc activity. As Tesco has often admitted it wasn’t until it began in-store picking for home delivery back in the 1990s and pickers came back with incomplete baskets that it realised the size of the problem.
Replenishment optimised
Today, of course, there are numerous IT tools that can help. Average sell-through rate, for example, can be tracked and highlighted if checkout records suggest a dip in performance. This is basically the solution adopted by French supermarket chain Monoprix. It is using an innovative replenishment system developed with NCR Teradata which monitors sell-through rates to first identify the top sellers and then identify when sales rates fall thus indicating out-of-stocks. From this a daily list of problem lines is generated for each store, alerting management to shortcomings. The lists are store-specific and reflect changes in consumer demand. The information is also being used to put pressure on suppliers to improve delivery schedules by highlighting just how much both trading partners could be losing in missed sales opportunities. Monoprix believes that return on investment can be achieved in less than a year.
‘We believed we had the potential to gain one per cent in sales across all departments by improving shelf availability,’ says Alain Carini, vp for supply chain at Monoprix, ‘But from our pilot programme this is likely to be much higher for food.’
Rather more sexy are the RFID and ‘smart shelf’ solutions regarded by some as the perfect panacea for stock-outs, although the technology remains very flaky. Ultimately tomorrow’s smart shelves would know when they were getting close to being empty.
In theory EPoS systems should give an accurate picture of what is being sold and, with broadband connectivity enabling real-time reporting, allow for continuous replenishment programmes so nothing should be out of stock for more than one delivery phase. In theory companies like Marks & Spencer should already be doing that but, as its high-profile RFID pilot item-tagging men’s clothing demonstrates, theory and practice do not always coincide. The M&S project involves a nightly high-speed stock-check using the RFID labels to drive overnight replenishment and has demonstrated increased sales and improved product availability. So why couldn’t the company do that with existing EPoS data giving item level sales information in real time?
The official answer talks of vast amounts of data and difficulties in extracting store and item-specific information quickly from regular polling routines but one also suspects a few other imponderables such as shrinkage and human error. After all, if a shop’s goods inwards record suggest that it has received three sky-blue-pink golf umbrellas but then one was donated to the local charity raffle, one was stolen and one had a damaged pack so was added to the ‘one-day-we’ll-get-round-to-sending-it-back-as-areturn’ heap in the stockroom, then there will be no records of any sales and these amazing umbrellas which could be must-haves for dozens of potential shoppers, will be delisted as unpopular.
Until that far-off day when we have universal RFID labelling, identifying what is in stock, what goes through the checkout and what is being pinched, keeping the shelves filled will have to depend on selective product tracking, accurate demand forecasts, and intelligent space management.
