The thesis is specific and uncomfortable: brands that treat grocery loyalty programs as a paid media channel are already losing, and the shift from points-and-perks to AI-driven personalisation will accelerate that loss faster than most commercial teams have planned for.
Tesco Clubcard, Kroger's loyalty ecosystem and Albertsons for U are not discount programs anymore. They are data platforms. The retailer's algorithm decides which offer reaches which shopper at which moment. Your brand does not control that decision. The retailer does. And the more you behave like a media buyer, writing a cheque for featured placement and measuring redemption rates, the less signal you generate that the algorithm can act on.
Why the Old Loyalty Model Is Breaking Down
The points-and-perks model was built for a world where the retailer's main job was to aggregate traffic and the brand's main job was to buy attention inside that traffic. Fund the promotion, clip the coupon, count the uplift. It worked when the retailer had limited data and the brand had the marketing budget.
That world is gone. A grocery loyalty program running on modern AI infrastructure can identify a shopper who is lapsing from your category three weeks before they stop buying. It can serve a personalised offer at the exact moment a competing brand runs a trial promotion. It can suppress a discount for a shopper who was going to buy anyway and redirect that spend toward someone genuinely at risk of switching. The retailer can do all of this. You cannot, unless you are working with the retailer rather than simply paying them.
The commercial pressure to act is real and immediate. Food Business News reports that a McKinsey analysis shows volume growth for the food and beverage sector now sits below 1 percent annually, far below historical norms. Since 2023, total shareholder return for major global CPG companies is down roughly 7 percent, while the broader S&P 500 is up 9 percent. That gap does not close by spending more on trade promotions structured around programs that are becoming obsolete.
What AI-Driven Personalisation Actually Changes
When a loyalty program runs on AI personalisation, the unit of value shifts from the promotion to the relationship. The retailer is no longer selling you an end-cap or a featured coupon slot. They are offering you access to a specific segment of shoppers at a moment of high purchase intent, shaped by behavioural data you do not own.
That changes three things for brands.
First, relevance beats reach. A personalised offer to 40,000 high-probability buyers is worth more than a mass coupon to 2 million Clubcard holders. But to build that offer, the retailer needs to know something about your brand that their transaction data alone does not tell them. What drives your shopper's repurchase? What occasions are growing? What does your innovation pipeline tell you about where their needs are going? If you share that, you become a partner. If you do not, you are a vendor.
Second, the algorithm rewards consistency. AI personalisation systems optimise over time. A brand that shows up with irregular promotions and inconsistent messaging trains the algorithm to treat it as noise. A brand with a coherent always-on presence, tied to real shopper need states, earns a structural advantage in how often it surfaces in personalised recommendations.
Third, data reciprocity is now a commercial negotiation. The retailers running sophisticated loyalty programs are beginning to offer brands something genuinely new: aggregated, anonymised insight into how their shoppers behave across the whole basket, not just in one category. That is more valuable than any panel study. But you have to bring something to the table in return. The brands that understand this are restructuring their customer marketing teams to include loyalty data specialists who speak the retailer's analytical language.
The Sceptic's Objection
The commercial director reading this will raise a fair objection: the retailer has always had more data than the brand, and yet trade promotion spending has grown for two decades. Why is this moment different?
The answer is that the data gap has crossed a threshold. A loyalty program that showed you a shopper's purchase history was useful but not transformative. A program that predicts future behaviour, identifies switching risk in real time, and serves personalised content across app, email and in-store digital screens is a different instrument entirely. The brand that is not plugged into that system is not just missing an opportunity. It is invisible to the moments that matter most.
There is also a structural signal in the CPG growth data. The McKinsey analysis reported by Food Business News shows that from 2002 to 2012, publicly held CPG companies averaged 9 percent annual money growth. That era ended. Price rises during the 2021 to 2022 inflation period masked the underlying problem, but the masks are coming off. Post Holdings' incoming CEO Nicolas Catoggio, speaking on the company's second-quarter earnings call, said that if inflation stays in the low single digits, CPG companies are more likely to absorb costs within their margins rather than raise prices, potentially by lowering how much they spend on promotions. If promotional spend contracts across the category, the brands with the strongest earned position inside retailer loyalty ecosystems will take share. The brands still renting attention will not.
What the New Participation Model Looks Like
Earning your place in an AI-driven loyalty program is not about spending more. It is about spending differently and building a different kind of relationship with the retailer's data and marketing teams.
The brands making this shift are doing four things that the old model never required.
They are sharing forward-looking insight, not just historical sales data. If your innovation team knows that a new protein format is launching in six months and it maps to a shopper need state the retailer's data has identified, that conversation belongs in your joint business plan, not just in your sales deck. Food Business News reports that smaller, more agile competitors have already gained ground on big CPG brands partly because they move faster on consumer need states. Loyalty data gives you the map. You need to act on it.
They are designing promotions for the algorithm, not the category manager. This means fewer broad discounts and more targeted offers tied to specific occasions, life stages or basket compositions the loyalty data has surfaced. A broad price cut trains the program to see your brand as a commodity. A personalised offer tied to a relevant occasion trains it to see you as a high-quality signal.
They are putting loyalty analytics inside the commercial team, not in a separate digital marketing silo. The brands losing ground in this transition are often the ones where the trade marketing team owns the retailer relationship and the digital team owns the loyalty conversation, and neither speaks to the other before the joint business plan is signed.
And they are measuring earned loyalty separately from bought loyalty. Repeat purchase rate among shoppers who received a personalised offer versus those who received a mass promotion is a KPI that most brands do not yet track. It should be the first number on your loyalty dashboard, because it is the number that tells you whether you are building something durable or just renting traffic.
What to Do Next Week
If you have a joint business plan review coming up with any major grocery account in the next 30 days, add one question to the agenda: what behavioural segments inside their loyalty program are most relevant to your category, and what would it take to build a co-funded personalised offer targeted at those shoppers specifically?
You are not asking for a briefing. You are proposing a data partnership. The retailer's loyalty team will engage with that differently from how they engage with a trade spend negotiation.
If you do not have a review soon, start internally. Map where your current loyalty-related spend sits and separate the bought from the earned. Bought is any spend that purchases placement or reach inside a loyalty program on a transactional basis. Earned is any activity that generates a recurring signal the algorithm can use to serve your brand to the right shopper at the right moment.
The gap between those two numbers is the gap between where you are and where you need to be. In a market where CPG volume growth is below 1 percent and the retailers have more data about your shoppers than you do, closing that gap is not a marketing initiative. It is a commercial priority.