Latest / The Retention Marketing Podcast with Fexingo: Email, SMS, and Loyalty for Repeat Customers / How Supermarkets Use Receipt Data to Drive Loyalty
Transcript
- Lucas: The modern loyalty card is dead. I know that sounds provocative, but look at the average supermarket receipt. You have your points balance, maybe a few coupons clipped digitally, and then a list of generic offers at the bottom that nobody reads. It’s noise. Luna: It feels like we’re being talked at rather than spoken with. If I buy diapers on Tuesday, why am I getting an offer for cereal on Friday when I haven’t bought cereal in months? Lucas: Exactly. The industry is stuck in a batch-processing mindset. They collect data, aggregate it into broad segments like 'millennial parents' or 'budget shoppers,' and blast static coupons. But real retention happens in the micro-moments. Take Kroger. They aren’t just tracking what you buy; they’re predicting what you’ll need next week based on your consumption velocity. Luna: Consumption velocity? So they know exactly when my milk runs out before I even realize it’s empty? Lucas: They know when you’re likely to run out, and they serve you a personalized discount on milk at the exact moment you scan your loyalty card at the self-checkout. It’s not a coupon you clipped three days ago. It’s a dynamic offer generated seconds before you pay. That’s the difference between a transaction and a relationship. Luna: That requires a level of data integration most retailers don’t have. How do they actually pull that off without it feeling creepy? Lucas: It stops feeling creepy when it works. Most people complain about privacy until they get a fifty-cent discount on eggs they were already planning to buy. The trick is relevance. If I’m buying cat food, I don’t want an offer for baby formula. I want an offer for the premium brand I usually skip due to price, nudged by a targeted discount. It respects my budget and my habits simultaneously. Luna: So it’s less about rewarding past behavior and more about influencing future intent. That’s a fundamental shift in how we measure ROI on loyalty programs. Lucas: Right. Traditional metrics look at redemption rates. Personalized offers look at incremental lift. Did this customer buy something they wouldn’t have otherwise? Or did they just switch brands because they got a better deal? The latter is dangerous if it erodes margin, but the former builds genuine stickiness. Luna: And how do you tell the difference? Because once you start giving everyone a different price, you risk alienating the people who don’t get the deal. Lucas: You control for it with A/B testing on a massive scale. Kroger tests these offers against control groups constantly. If the personalized group doesn’t show higher lifetime value over six months, the algorithm kills that segment. It’s brutal efficiency. But there’s a catch. It relies heavily on first-party data, which is becoming harder to get as third-party cookies die and privacy laws tighten. Luna: So the players who win here are the ones who can convince customers to hand over their receipt data voluntarily. That’s a trust game now. Lucas: Absolutely. The value exchange has to be transparent. You give them your purchase history; they give you a better price. If they sell that data to advertisers, the contract breaks. The successful models keep the insights internal, using them only to optimize inventory and pricing. It’s a closed loop. Luna: I wonder if this model translates outside of groceries. Does a boutique coffee shop ever have the data volume to do this kind of prediction? Lucas: Not at that granularity, no. But the principle applies. Small businesses can use simpler versions. If a customer buys a specific roast every two weeks, a simple automated email saying 'Your usual is back in stock' is a primitive form of this. It’s not ai driven, but it’s personalized. The gap is just the sophistication of the tooling. Luna: So the barrier to entry isn’t really the technology anymore. It’s the willingness to change the culture from 'collect points' to 'solve problems.' Lucas: Precisely. Most loyalty programs are designed by finance teams to manage liability. They want to minimize the cost of the reward. Marketing teams should be designing them to maximize engagement. When those two goals collide, you get the stale, one-size-fits-all programs we see everywhere. Luna: That’s a great distinction. Finance sees points as a debt on the balance sheet. Marketing sees them as a ticket to conversation. Which one wins in practice? Lucas: Usually finance, unfortunately. Until the CMO proves that personalized offers drive higher margins. Once they show that targeted discounts bring in new categories of buyers, the pressure shifts. But it takes time. Most companies don’t have the patience for that multi-year horizon. Luna: Speaking of patience, how long does it take for a retailer to build this kind of infrastructure from scratch? Is it a five-year project? Lucas: For legacy grocers, it’s been twenty years of accumulation. For a new player today, it’s much faster. You can plug into platforms that already handle the predictive modeling. The hard part isn’t the code; it’s the clean data. If your POS system is fragmented across hundreds of stores, cleaning that up is a nightmare. Luna: So data hygiene is the hidden moat. Companies with messy databases are doomed to stay generic. Lucas: Doomed is strong, but stuck is accurate. They’ll survive on habit, but they won’t grow. Growth comes from understanding your customer better than they understand themselves. That’s the endgame. Luna: Do you think this will lead to a bifurcation in retail? Where some brands feel incredibly smart and others feel like spam filters? Lucas: It already has. Think about your own apps. You probably ignore half of them immediately. The ones that stay active are the ones that seem to anticipate your needs. Amazon Prime is the extreme version of this. It’s not just fast shipping; it’s the recommendation engine that makes finding products effortless. That’s retention through friction removal. Luna: Friction removal is interesting because usually we talk about adding value. Here, removing the cognitive load of decision-making is the value proposition. Lucas: Exactly. In an economy where inflation is squeezing disposable income, every minute spent comparing prices is a tax on the consumer. If a brand can save you that time AND money, you become loyal out of necessity, not just affection. And necessity-based loyalty is very durable. Luna: Necessity-based loyalty. I like that phrase. It sounds less fluffy than 'brand love.' It’s pragmatic. Do you think small businesses can compete on pragmatism without big data? Lucas: They have to. They can’t beat the algorithms, but they can beat the anonymity. A local bakery remembering your order is a primitive algorithm, but it’s human. The tech should augment that, not replace it. If a small business uses a simple CRM to note preferences, that’s enough to create that sense of being known. Luna: So the takeaway is: stop buying points. Start buying insight. Invest in knowing your customer, not just counting their transactions. Lucas: Spot on. If you’re spending budget on a points system, ask yourself: does this help me understand why they left? Or does it just make them stay longer while they figure out how to leave anyway? Luna: Ouch. That’s a hard question to answer honestly. But necessary. Lucas: It is. And that leads us to how we keep these conversations going. We deliberately don't run ads on these shows because we want to focus on substance over clicks. If these marketing conversations have sparked something you've actually used or found useful in your work, the link is buy me a coffee dot com slash fexingo. Luna: It’s a small way to keep the signal to noise ratio high for both of us, and for you. Lucas: Appreciate it. Now, back to the data. One last thing on the grocery example. What happens when the algorithm gets it wrong? Like if it keeps offering steak to someone who went vegan? Luna: That’s the failure mode. It feels invasive. You’d probably delete the app or stop using the card immediately. Trust is fragile. Lucas: Agreed. The exit ramp must be easy. If you can’t opt out of the personalization easily, you lose the customer entirely. It’s a delicate balance between helpful and haunting.