The structural shift that changes everything
Loyalty programmes have always been about retention. What is new in 2026 is that the best ones are also about precision. Tesco Clubcard, Kroger Boost, and Albertsons for U are no longer running points schemes with the occasional personalised coupon layered on top. They are running AI-driven personalisation engines that decide, in real time, what offer, what product, and what message reaches each individual household.
That distinction matters enormously for brands. A programme built on blanket discounts can be matched by any competitor willing to price lower. A programme built on knowing that a specific household shops for high-fibre snacks, avoids artificial colours, and typically buys on Thursday evenings is far harder to displace. The retailer that holds that data sits at the centre of the shopper relationship in a way that no brand can replicate from the outside.
P2PI's framing captures what is happening: the next era of loyalty will not be bought. It will be earned. And the mechanism for earning it is relevance, not generosity.
Where the three reference programmes stand today
Tesco Clubcard remains the most cited example of loyalty data used as a commercial asset. The programme gives Tesco a level of household-level purchase visibility that it monetises both through personalised offers and through its retail media network, Tesco Media and Insight Platform. For brands, Clubcard data is increasingly the lens through which ranging and promotional decisions are made. Getting access to that data, or building joint plans that are anchored to it, is now a core competence for any supplier with Tesco as a key account.
Kroger's Boost membership tier sits above its standard loyalty scheme and bundles fuel savings, free delivery, and personalised offers into a single paid subscription. The paid tier creates a self-selecting group of high-frequency, high-basket shoppers, exactly the audience brands most want to reach. Kroger's 84.51 analytics subsidiary converts that data into targeted media and promotional recommendations, meaning the brand's promotional plan and the retailer's loyalty engine are increasingly operating on the same signal.
Albertsons for U takes a similar approach with its AI-driven offer personalisation. The programme uses purchase history and behavioural signals to surface offers that are specific to the individual shopper rather than to a broad segment. For brands, the implication is the same: a category-level funded promotion delivered through the programme will be less effective than an offer designed around the specific household behaviours the programme can identify.
Why the old funding model is losing ground
The classic loyalty funding model works like this: a brand funds a price promotion, the retailer pushes it to a broad loyalty audience, and both sides measure the sales lift. That model is not disappearing overnight, but it is losing structural ground for two reasons.
First, the algorithm has more information than the funded promotion. If the retailer's AI already knows that a household never buys a particular brand category, showing them a discount on that brand generates noise, not conversion. The programme's credibility with the shopper depends on relevance, and irrelevant offers erode that credibility over time.
Second, private label is the default beneficiary when brand offers are not competitive. As prior analysis in this series has noted, the retailers building these AI personalisation engines are also the retailers with the strongest own-label programmes. When a brand's offer loses in the algorithm's ranking, the slot does not go empty. It goes to the retailer's own product.
The McKinsey analysis cited in earlier coverage of this topic showed CPG volume growth sitting below 1 percent annually and total shareholder returns for major global players down roughly 7 percent since 2023. In that environment, loyalty spend that does not convert is not just wasteful. It actively subsidises the retailer's case for ranging its own label instead.
What AI personalisation means for promotional planning
The shift to AI-driven personalisation changes the unit of planning. The old unit was the promotion: a depth, a duration, a mechanic, a funded cost. The new unit is the household signal: what does this shopper buy, when, in what combination, and what would move them?
For brands, three practical changes follow from that.
First, the data you bring to the joint business plan matters as much as the money you bring. Retailers running AI personalisation programmes want brand partners who can contribute consumer insight, not just promotional budget. If you can tell Kroger's 84.51 team something about your buyer's purchase triggers that they cannot see in their own transaction data, you are a more valuable partner.
Second, the media and loyalty budgets need to talk to each other. Retail media networks and loyalty personalisation programmes are converging fast. Tesco's Clubcard data feeds its media platform. Kroger's 84.51 analytics feeds both Boost offers and its retail media network. If your retail media plan and your loyalty investment plan are being run by different teams with different KPIs, you are missing the compounding effect of the two working together.
Third, the measurement conversation needs to shift from promotional sales lift to household-level conversion. A blanket funded promotion can show a sales lift number. An AI-personalised offer can show you which household types responded, what they bought alongside it, and whether they came back. That second set of answers is far more useful for building a plan that earns loyalty rather than renting it.
What to watch next
Two signals are worth tracking closely in the second half of 2026.
The first is how retailers price access to their AI personalisation tiers for brand partners. Tesco, Kroger, and Albertsons have all built the infrastructure. The commercial question is how they package access to it. Expect clearer tiering of brand partnerships based on data contribution and media spend, not just promotional funding.
The second is the role of AI in brand-side personalisation. As signals S17 and S20 from this brief period show, major food manufacturers including Nestlé, Coca-Cola, Diageo, Mondelez, PepsiCo, and Mars are deploying AI across their innovation and marketing workflows. The brands that extend that capability into shopper marketing, using AI to build offers that compete with the retailer's own algorithm on relevance, will be better placed to earn their way into the programme rather than buying their way in.
The P2PI framing is right. Loyalty will not be bought. But earned loyalty still requires investment. The investment just needs to be in precision, data, and relevance rather than in discount depth. For commercial leaders heading into H2 planning cycles, that is the reframe worth making now.