Leading Supermarket retailers knows a great deal about their customers. A loyalty account reveals what a household buys and how often it buys it. A mobile app can show what a shopper searches for, which coupons are clipped, and which products are placed in a cart and then abandoned. An online order adds location, delivery history, and browsing behavior to the picture.
Most retailers use this information to market more effectively. The customer gets a coupon. The retailer gets a better chance of making a sale.
The difficult question begins when customer information helps determine the price itself.
That is the concern behind what regulators, researchers, and privacy advocates call surveillance pricing. The term is dramatic, but the basic idea is simple: a business collects information about a person, makes a judgment about that person’s willingness to pay, and uses that judgment to influence the price or discount offered.
The subject is often discussed as if every different price must be surveillance pricing. That is not right. Grocery prices differ for many ordinary reasons. The important work is to separate those reasons from the use of personal data.
Four different kinds of price difference
Start with four examples.
- A retailer changes the price of olive oil because its supplier raised the wholesale price. That is a normal business response.
- A retailer changes a price in one store because the cost of labor, rent, or distribution is different there. That is a location-based price difference, not necessarily a personal one.
- A loyalty member receives a coupon for coffee because the retailer knows that customer buys coffee regularly. That is a personalized promotion.
- A shopper sees a higher price because an algorithm predicts that shopper is less likely to compare prices or switch brands. That is much closer to surveillance pricing.
There is also a fifth case that can confuse the discussion: an online retailer or delivery platform may test different prices with randomly selected groups of shoppers. That is a pricing experiment. It may be unfair or poorly disclosed, but it is not automatically personalized pricing if the customer’s personal profile did not determine the result.
The distinction can be reduced to two questions:
Did the price change because the market or the retailer’s circumstances changed? Or did it change because the system made a judgment about this particular customer?
That is the line industry and policymakers need to understand.
Loyalty programs are where the issue starts
The trade-off behind a loyalty program seems straightforward. The shopper identifies themselves and receives a discount. The retailer receives information about the shopper’s behavior.
The information collected behind the account can be much broader than the transaction at the checkout counter. Retailer disclosures may cover purchase history, location, online activity, device information, and inferences about a customer or household.
That does not mean the retailer is using all of that information to set the base price. In many cases, it is being used to decide which promotion to show, which product to recommend, or which advertisement to serve.
Consider a simple example. A jar of coffee has a posted price of $8. Every customer sees the same $8 price, but a frequent coffee buyer receives a $2 coupon. The customer gets a targeted promotion; the shelf price is uniform.
Now change the example. One shopper sees $8 and another sees $9 because the system predicts that the second shopper has fewer alternatives or is less likely to shop around. The underlying price has become personal.
That difference matters because consumers generally understand that a loyalty program offers conditional discounts. They may not understand that a retailer is estimating their income, financial pressure, brand loyalty, or willingness to pay and using that estimate to set a price.
An inaccurate promotion profile is annoying. An inaccurate pricing profile is more serious, particularly if the shopper has no way to see or challenge the inference.
The grocery receipt is also a data record
For a grocery retailer, the receipt is more than a record of a completed sale. Over time, it becomes a detailed picture of a household’s routines.
The retailer may know which coffee a customer buys, how frequently it is purchased, whether the customer chooses organic products, which coupons are redeemed, which store is visited, and what happens when prices change.
That information is valuable even if it never affects the base price. It can support advertising, assortment decisions, demand forecasting, promotion design, and supplier-funded marketing programs.
This is why the surveillance-pricing question cannot be separated completely from retail media and loyalty economics. The same data can support several commercial activities. A customer may willingly exchange information for a discount, but the terms of that exchange are not always clear.
The practical question for a retailer is not simply, “Are we using customer data?” The answer will usually be yes. The better questions are:
- Which data is being used for marketing?
- Which data is being used for promotions?
- Which data is being used for pricing?
- What information is inferred rather than directly provided?
- Can the retailer explain the difference to a reasonable customer?
What digital grocery platforms have shown
Online grocery platforms have made the issue more visible because they can display different offers and prices without putting different paper tags on the same shelf.
Investigations into online grocery pricing have found instances in which the same products from the same retailers appeared at different prices for different shoppers. In one widely reported investigation, some differences reached 23%.[1]
The platform’s explanation was that the tests were randomized rather than based on personal information. If that explanation is accurate, the finding is not proof that the platform used a personal profile to price each customer. It is evidence of something slightly different but still important: consumers may be placed into pricing experiments without understanding that an experiment is taking place.
From the shopper’s point of view, the screen does not explain the difference. The price may reflect a promotion, inventory, delivery economics, a competitor’s price, a test group, or a profile-based decision. The customer sees the number but not the reasoning behind it.
That opacity is the common thread running through the entire debate.
Electronic shelf labels are not the same thing
Electronic shelf labels have also attracted attention because they make price changes easier. A retailer can update thousands of labels through software instead of sending employees through the store with paper tags.
The technology itself is neutral. It can reduce labor, improve accuracy, and make promotions easier to manage. It can also make frequent price changes technically easier.
But a digital label does not prove that a retailer is tracking the person standing in front of it. The important question is what pricing rule sits behind the label.
A price that changes because inventory is low is one thing. A price that changes because the system identifies a particular shopper and estimates that shopper’s willingness to pay is another.
Industry discussions become less useful when they treat the technology as the offense. The real issue is the combination of data, pricing rules, and disclosure.
Why retailers are interested
The economic attraction is obvious. A uniform price is simple, but it may leave money on the table. One shopper might have paid more; another might only buy if the price is lower.
Data-driven systems promise to narrow that gap. They may also help retailers make better offers, reduce waste, respond to inventory conditions, and compete more effectively. Those benefits are not imaginary. Personalization can produce a lower price for some shoppers, and targeted promotions can help a retailer move products or retain customers.
The concern is distribution. Who gets the lower price? Who gets the higher price? What information determined the result? Can the customer understand the rule? Can the customer go somewhere else? The answer depends on the market, the competitive alternatives, the accuracy of the data, and the retailer’s willingness to be transparent. There is no single answer for every algorithmic pricing system.
The questions that deserve attention
For food retailers, the most useful internal review is practical rather than theoretical:
- Map the data. Identify the personal, inferred, purchased, and operational data that enters pricing and promotion systems.
- Map the decisions. Document whether each data element affects assortment, advertising, coupons, delivery charges, the base price, or something else.
- Separate promotion from price. Make clear when a discount is conditional and available to other shoppers who meet the same published condition.
- Test accuracy. Determine whether important customer inferences can be wrong and whether a customer has any meaningful way to correct them.
- Explain the rule. Give shoppers a usable explanation when personal information affects a price or offer.
- Govern experiments. Set boundaries for randomized price tests, document approvals, and consider whether the practice would surprise a reasonable customer.
- Watch the law. Requirements differ by jurisdiction and continue to develop. Legal and compliance teams should review the current rules before a pricing system is launched.
For suppliers and technology companies, the same questions apply. A pricing or personalization product should specify what data it uses, what decisions it supports, how results are tested, and what disclosures the retailer needs to make.
The line worth watching
Different prices are not new. Grocery retailers have always adjusted prices for cost, location, inventory, competition, promotions, and customer programs.
What is new is the amount of information available to make those decisions—and the growing ability to process that information automatically.
The central question is therefore not whether grocery prices will change. They will.
The question is whether a shopper receives a different price because of a legitimate market condition, a clearly explained loyalty rule, a controlled experiment, or a private prediction about that shopper’s willingness to pay.
That distinction will shape the future of grocery retail. It deserves less headline heat and more careful attention from the people who design the systems, sell the technology, regulate the market, and buy the food.
For more on surveillance pricing Go-To: