How supermarkets are using AI across the store from ads to catching shoplifters – but is it making you spend more?
ROBOTS may not be packing your bags at the checkout – yet – but artificial intelligence is already being used by every British supermarket.
Retailers employ AI behind the scenes in all sorts of different ways, from helping to reduce food waste to spotting shoplifters and suggesting what products you should buy.
But is it really making our lives easier. Or is it pushing us to spend more money? Rosie Taylor looks at the pros and cons.
PREDICTING TRENDS
SUPERMARKETS have always tried to predict shopping trends to help them set prices.
For example, they might raise the costs of travel toiletries before the summer holidays, then slash them in the autumn.
AI allows them to analyse more information than ever before to accurately predict demand in each individual store.
Data can include local weather forecasts, traffic problems, the habits of shoppers in the area and even how customers move through the store.
For example, the retail insights provider Reapp has used AI to predict shoppers will buy 6.7million extra packs of beers during the Euros and 1.2million more bags of crisps.
Stores can use this data to change prices. They might make booze cheaper as they know shoppers are looking for the best deals.
But they may also up the cost of crisps, and move them to a shelf beside the beer, as data shows customers will be more likely to get pack if it’s convenient.
“Supermarkets will say they’re making customers’ lives easier but it also gives them more opportunities to make money,” says Matt Wills, co-founder of Acumen, a company which advises suppliers on how to price and promote products for supermarkets.
“In the past, supermarkets would have had to pay a market researcher to learn about customer behaviour. Now, with AI, they can get insights faster and in greater depth for a fraction of the cost.”
“It means people who are a more vulnerable demographic, could then miss a discount they would benefit from,” Matt adds.
CHANGING PRICE LABELS
ELECTRONIC labels on shelves make it much easier for supermarkets to change prices multiple times in a day.
Morrisons, Asda, Sainsbury’s and Tesco have all trialled electronic shelf labels, and Lidl quietly put them in stores earlier this year.
Lidl’s look similar to paper labels and its own research found two thirds of shoppers didn’t notice the “subtle change”.
But consumers may struggle to keep track of rapidly changing prices, warns Matt Wills.
He says: “The risk for consumers is fluctuating prices mean price points become much harder to remember, so it is hard to know whether you are getting value for money.”
Digital labels also make it easier for supermarkets to up prices and maximise profits if there is a surge in demand.
AI will track whether or not demand declines if supermarkets sell items at a higher price
Matt Wills
And Matt is concerned AI-powered variable pricing could penalise the most vulnerable customers, including those who aren’t able to shop around as they lack transport.
“AI will track whether or not demand declines if supermarkets sell items at a higher price,” he explains.
Such a system may notice shoppers in a town with lots of elderly residents always buy a lot of products, such as health supplements, regardless of cost.
It means the system will not reduce the price on these items in that town’s store, even if they are on offer in other outlets.
CUTTING WASTE
BY using AI to monitor stock levels and expiry dates, stores can drop prices to shift produce that will soon be past its best — like an automatic “yellow sticker” system.
Experts believe this could help cut the 300,000 tonnes of food supermarkets waste every year.
Peter Pugh-Jones, of AI-powered data firm Confluent, helps Sainsbury’s and other major supermarkets be more efficient.
He says AI is used in several ways to ensure customers get quality produce, from technology that checks how fresh items are while they are in transit to systems that cross-check expiry dates against expected arrival times of lorries, taking into account issues such as local traffic and weather.
He explains: “These systems can detect whether produce may spoil sooner than expected, and could even divert delivery vehicles to closer locations.
“It’s better for the environment if the produce arrives in store when it is still usable by consumers.”
TARGETED ADS
RETAILERS get data on customers’ needs from sales, loyalty card information and even smart cameras that monitor which items shoppers pick up but put back.
They use it to create tailored offers for online customers. So if you add certain products to your basket, an AI-powered system will suggest other goods.
Peter Pugh-Jones says: “Most of the time, AI is used without you realising, to help you do something faster or give you extra information, like suggesting products.”
Some retailers are working on systems that encourage shoppers to buy healthier items if they change their settings to show they are on a diet, he adds.
Algolia, used by supermarkets such as Co-op, uses AI in online searches to understand what shoppers are looking for, even if they key in the wrong words or typos.
So if you search for “gluten”, the system understands you probably mean gluten-free products.
The Tesco app also suggests extra items customers can buy.
Piyush Patel, of Algolia, says: “If a customer searches for pasta, the search might also suggest related items like pasta sauce or cheese, helping complete their meals.”
As a result, Co-op has seen a 45 per cent rise in the number of “additional items” online shoppers add to their baskets before paying.
But Matt Wills warns this may push customers to spend more than they can afford.
CLEVER CAMERAS
SUPERMARKETS are also using smart cameras to track and improve store stock and safety.
Morrisons has recently put AI-powered cameras in its stores to help staff fill shelves more quickly. Sainsbury’s is also introducing a similar system.
The cameras monitor shelves and detect when products are in the wrong place or running low, so staff can prioritise the areas needing attention instead of looking around the store.
Although supermarkets say this will free up employees to spend more time on customer service, critics fear it could lead to job cuts.
Tesco, Asda and Co-op have been trialling technology which scans customers’ faces at self-checkouts and assesses their age when they try to buy age-restricted items such as alcohol.
If the system picks up that a customer appears to be under 25, they are asked for ID.
But it saves wasting staff and shopper time by targeting everyone.
Sainsbury’s has used AI cameras to detect if someone puts an item in their pocket.
If it spots suspicious activity, it alerts security. The store said it helped stop 6,000 theft attempts in six months between 2019 and 2020.
Several big retailers, including convenience store chain Nisa, are also using facial recognition technology to catch shoplifters.
But there are reports of customers complaining they have been wrongly profiled as criminals by the technology.
AI has the potential to benefit customers, says Matt Wills, but it must be well managed. He adds: “We need to ensure people aren’t discriminated against.”