The Rise of Personalised Pricing in Grocery

Speaker A: By the end of this episode, you will hear how one of Britain's best known high street chains could tell from loyalty data alone when a shopper was trying to give up smoking. And exactly what it did. With that knowledge, the single shelf price, the number that has anchored grocery for a century, is being quietly dismantled. And the most consequential pricing decisions in our industry are, uh, no longer taken at the shelf. They are taken one shopper at a time, inside an app. Welcome to FMCG Weekly, brought to you by Acurus.com, the leading independent consultancy for revenue growth management. I am Matilda, your AI host. There is a lot of loose talk about pricing, so fix on the one distinction that matters. Dynamic pricing moves a price for everyone at once in response to demand. We have all accepted that the airline seat beside us cost a different fare. Personalized pricing is something else. It sets a unique price or net promotion for a single named individual for the same product at the same moment, not to manage stock, but to estimate and capture each shopper's willingness to pay. In plain terms, the capture of consumer surplus, one person at a time. Food, though, is not an airline seat. Groceries are essential and visible, and a supermarket that openly surged a shelf price would face a backlash no margin could justify. So the industry has made a quiet choice. Rather than move the visible price, it routes personalization through the one channel. Shoppers have opted the loyalty scheme. The result is dual pricing, where two shoppers at the same shelf are offered two different net prices and only one of them can see both. Nowhere is this further advanced than the United Kingdom. Tesco set the template with club card prices and Sainsbury's followed with nectar prices. But the structural move is subtle. Standard discounts have been withdrawn from non members, so the shelf price a non member sees is effectively a penalty. To avoid it, you register and to register, you hand over your data. The loyalty card has become the entry ticket to ordinary value. From there, it deepens. Tesco has trialed a hyper personalized tier of club card prices drawn from each shopper's own history and partnered with Adobe to embed agentic artificial intelligence that anticipates what a shopper will want next and generates the offer in real time inside the app. To see how granular this gets, consider Boots, which used its Advantage card data to identify shoppers who appeared to be giving up smoking and served them marketing for nicotine replacement. That is the level of inference now in play. Not a segment, but a specific person and a specific intention. The Competition authority has given the model A qualified pass across some 50,000 products over a year. It found about 92% of loyalty promotions were genuine savings with member discounts of 17 to 25%. But two findings deserved more attention than they got. First, exclusion. These benefits are largely unavailable to the elderly, to under 18s and to anyone without a smartphone, opening a value gap between the connected and the unconnected. Second, and more startling, the bank of England has had to fold person specific discounts into the way it measures national inflation. Because when every household pays a personalized price, the idea of a single representative price begins to dissolve. When the official inflation measure has to be rebuilt around personalization, this is no longer a tactic, it is structural.

Speaker B: Is your company selling products in a mature category? There is almost no new demand left to create. The growth you are chasing already exists. It is just sitting in someone else's basket, or in a less valuable product, or in the wrong store. Growth is not something you generate, it is something you steershoppers, switching towards your brands, towards your more valuable products, towards your more valuable customers. Acurus, uh, helps you do exactly that. The analysis to see where your growth can genuinely come from and the plans and forecasts to steer it there. Find out how@acurus.com that's accuris.com welcome back.

Speaker A: Europe shows there is no single playbook. In the Netherlands, Albert Hein has aimed its algorithms not at people, but at perishables, letting the price of fresh poultry and fish fall automatically every 15 minutes as it nears expiry, protecting margin on stock that would otherwise be written off and cutting food waste by more than 250 tons a year. Germany shows the opposite obsession ownership of data. At the end of 2024, Riwe walked away from a 14 year partnership with the Payback Coalition and within a day launched its own scheme, paying Eurocredit rather than points, with a spend escalator rising to a 10% reward, all to keep its customer data for its own media and delivery business. Tellingly, Ithaca bet the other way, joining Payback to unify its merchants and found the promised sales lift largely failed to appear. The architecture of the program and who owns the data beneath it now matters more than the generosity of the rewards. And in France, Carrefour is spending around 100 million euros a year on artificial intelligence, steering personalized discounts onto its own private label so it can show a sharp net price while protecting its margin for suppliers. The uncomfortable truth is that the moment of pricing is migrating out of your control, when the decisive event is a personalized coupon inside a retailer's app execution sits on an interface the retailer owns and you cannot see the strongest responses are concrete. Nestle matched its brands against retailer loyalty profiles with with the data firm Dunhumby and reported an 11% uplift at the household level. And this May, Coca Cola, Unilever, Nestle, PepsiCo and Bayer joined with Kantar to build a single measure of advertising impact so that as retailers push them to spend in proprietary media, they can audit whether it actually works. The most instructive case, though, is Coca Cola's refusal to personalize at all. It abandoned an experiment decades ago with a vending machine that raised prices in hot weather, and the logic still holds because it sells concentrate to independent bottlers. Uniform pricing stabilizes their economics and stops powerful retailers grinding down its inputs. Sometimes uniformity is the strategic asset. Remember too, that the growth the giants posted in 2023 was almost entirely a pricing story. High single digit organic growth built on pricing of 7 to 10 points while volumes were flat or falling. That move is spent broad list price rises are no longer available to price aware shoppers, so what growth remains has to be found surgically. The blunt lever has been replaced by a scalpel, which is why personalized loyalty shielded mechanics have become central. None of this is happening quietly in Europe. Any advertised discount must now be measured against the lowest price in the previous 30 days. Retailers must disclose when a price has been personalized by automated profiling, and penalties reach 4% of global turnover. In the United States, the fight has a surveillance pricing Maryland is the first state to ban grocers and delivery services from from using personal data to raise prices. More than 70 related bills have been tabled this year. And the case that lit the fuse Several hundred shoppers loaded identical baskets at the same online grocer simultaneously and found different prices on three quarters of the products, some varying by as much as 23%. The platform Instacart called it randomized testing and has stopped it. And regulators are already past disclosure onto a harder fear that independent pricing algorithms, each watching the others, could learn to hold prices above competitive levels with no human ever agreeing to collude. Expect auditability and explainability to become compliance requirements, not virtues. Underneath all of it sits trust. Around two thirds of British consumers dislike dynamic pricing and and a third actively detest it. And the root is information asymmetry. The algorithm sees data the shopper cannot. There is a double edge here that is easy to miss. The same system that charges a desperate shopper more can hand a hesitant one a discount, which means a shopper who blocks all tracking may simply never see the offer and will only ever be shown the list price, the highest price on the table. In this market, privacy is not free. That is it for this week's episode of FMCG M Weekly. Thank you for making it to the end and see you next week.