The Commission’s proposed enforcement policy is being framed largely around predatory retail pricing. Its treatment of individualized discounts, consumer data and disclosure raises important questions for loyalty, rewards and incentives.
The Federal Trade Commission is considering an enforcement policy that would require companies to tell consumers when the price they are seeing has been personalized using information about them. Most of the attention around the proposal has focused on the obvious concern: retailers using increasingly sophisticated data and technology to figure out how much an individual consumer is willing to pay, and then charging accordingly.
This isn’t just about personalized pricing.
Predatory pricing is very much part of what concerns the FTC. But a closer reading of the Commission’s proposed enforcement policy statement reveals a broader issue: the use of information about an individual consumer to determine the economic proposition presented to that consumer.
And price is only one way to do that.
What the FTC is actually proposing
The FTC is not proposing to ban personalized pricing. In fact, the Commission explicitly acknowledges that Congress has not given it authority to prohibit the practice categorically. Instead, the FTC is approaching personalized pricing through Section 5 of the FTC Act, which prohibits unfair or deceptive acts or practices.
The Commission’s argument starts with consumer expectations. In many markets, consumers reasonably assume that the price they see is generally the price other consumers would see under the same circumstances. There are obvious exceptions. Consumers understand that insurance prices reflect individual risk, that airline fares change and that ride-sharing prices can respond to local supply and demand. The FTC itself recognizes these distinctions.
But when a consumer sees a product priced at $12.99 in a retail store or on a website, she generally doesn’t assume that another customer might see $10.99 because an algorithm has concluded that he is more price-sensitive.
Technology is making that kind of distinction increasingly possible. The FTC points to the extraordinary amount of information generated and collected about consumers and the ability to use that information to infer things like willingness to pay or propensity to comparison shop.
The proposed policy moves the discussion from whether that capability exists to what businesses should be required to tell consumers when they use it. Where consumers reasonably expect prices not to vary based on their personal information, the FTC argues that failing to disclose personalization could constitute deception.
More significantly, the contemplated disclosure goes well beyond a generic statement that “prices may vary.” The Commission says businesses should clearly disclose that a price is personalized, the basis for the personalization and the types of data on which it is based.
The consumer isn’t merely being told that prices can change. The consumer is potentially being told that something about them changed the economics of the transaction.
From personalized pricing to personalized value
Price is only one mechanism through which companies alter the economics of a transaction. Consider a product with a $100 price. If an algorithm determines that one consumer is willing to pay $110 and presents that consumer with a $110 price, we have a straightforward example of personalized pricing.
But suppose the price remains $100. One consumer receives a personalized $20 discount. Another receives $5. A third receives nothing. Or everyone pays $100, but one consumer receives $20 cashback. Or 5X points. Or a $20 statement credit. Or a merchant-funded reward.
The sticker price hasn’t changed. The economics of the transaction have.
That creates a less obvious boundary: When does personalized promotion, reward or incentive become functionally equivalent to personalized pricing?
The FTC’s proposed statement does not answer that question, nor does it suggest that loyalty programs or personalized rewards should automatically be treated as personalized pricing.
But it does already venture into some of this territory. In one example, the FTC describes a consumer who believes a personalized price is a discount based on purchase history with a retailer when it is actually a higher price based on information about disposable income or shopping behavior elsewhere. The potential deception, in the FTC’s analysis, lies not simply in what the consumer ultimately pays but in what the consumer understands about why that economic treatment was offered.
That example matters because the FTC is already looking beyond the nominal mechanism — price or discount — to the data and decisioning that produced the economic outcome.
And that matters for the incentives industry.
A personalized pricing system might attempt to determine the most a particular customer will pay while maintaining an acceptable probability that she completes the transaction. A personalized incentive system might attempt to determine the least additional value the company needs to give that same customer to produce an acceptable probability that she completes the transaction.
Those are different questions. In some respects, they are opposite questions. But increasingly they can be answered using the same customer data, behavioral signals, predictive modeling and AI-driven decisioning technology.
Both ultimately affect the economic proposition presented to a particular individual.
The incentives industry has been moving toward this for years
The loyalty and incentives business has spent the better part of two decades trying to become more personalized. Mass offers became segmented offers. Segments became microsegments. Microsegments became individual targeting. Increasingly, AI makes it possible for individualized decisions to occur dynamically and in real time.
The objective is no longer simply to determine which customers should receive a promotion. Sophisticated systems can determine who should receive value, what form that value should take, how much should be offered, when it should be delivered and what behavior should trigger it.
Taken to its logical conclusion, the optimization problem becomes fairly simple: What is the minimum amount of incremental value required to change this particular customer’s behavior?
Traditional loyalty programs are probably not the interesting regulatory problem here. Spend $1 and earn one point. Reach Gold status and earn 2X. Members receive 10% off on Tuesdays. Those are transparent program rules. Consumers understand the proposition, and customers satisfying the same criteria generally receive the same treatment.
Increasingly sophisticated incentive programs don’t necessarily operate that way. Customer A may receive $25. Customer B receives $10. Customer C receives nothing. Not because of a published tier or clearly stated program rule, but because a decisioning system has determined that those are the optimal amounts required to produce a particular behavior from each customer.
The industry calls that personalization.
Through another lens, it is algorithmic allocation of customer value.
Transparency changes the equation
Nothing in the FTC’s current proposal suggests that the Commission intends to prohibit personalized rewards or loyalty incentives. But its treatment of transparency raises questions that overlap directly with increasingly sophisticated incentive practices.
The Commission’s position is that disclosure of personalization alone may not be sufficient. Businesses should disclose the basis for personalized pricing and the types of data involved. The FTC even says that describing something merely as a “specially selected” price would likely be misleading if important information about how the price was determined were omitted.
Apply that thinking to a modern incentive engine.
Imagine telling a consumer that a $15 offer was selected based on purchase history, browsing activity, estimated propensity to purchase and predicted likelihood of attrition. That may be a reasonably accurate description of how a sophisticated personalized offer is generated. It also sounds very different from “an exclusive offer selected just for you.”
Personalization has generally been presented as a customer benefit. Often it is. Better data can allow brands to make offers more relevant and put promotional dollars where they actually create incremental value.
But personalization also means deciding who gets more, who gets less and who gets nothing.
Suppose one customer receives a $25 incentive because a model determines she is unlikely to buy without it. Another receives $10 because the model thinks that will be sufficient. A third receives nothing because the model predicts he will buy anyway. All three decisions are economically rational.
Now suppose inferred income, browsing behavior or third-party data tells the model that one customer is less price-sensitive and therefore requires a smaller incentive. The company is still providing a benefit. But information about the consumer is now producing materially different economic treatment.
That is where “it’s a reward, not a price” becomes an incomplete answer.
Where the line may actually matter
Personalized promotions are hardly new. What is changing is their precision, scale and opacity.
Traditional segmentation might determine that customers fitting a particular profile receive a 20% offer. That rule can be documented, explained and audited. An AI-driven system can theoretically determine that a particular customer should receive $13.72 because that amount maximizes the expected incremental return from that individual.
There may eventually be no “offer strategy” in the traditional sense. The offer is the model output.
A system designed to determine the highest price a consumer will tolerate and one designed to determine the smallest incentive required to produce a transaction can have very different effects on consumer welfare. That distinction matters. Personalized incentives can create real consumer value, and treating every individualized benefit as equivalent to personalized pricing would ignore that difference.
But whether value flows toward or away from the consumer cannot be the only consideration. Personalized benefits can themselves create substantially different effective transaction economics.
A more useful way to examine these practices is to look at several things together: the direction of the value, the basis for the differentiation, the information used to make the decision, the transparency of the rules and what a consumer could reasonably expect.
A published loyalty program in which members earn benefits by satisfying known criteria looks relatively straightforward. The consumer understands why the value differs. A targeted promotion based on previous purchase behavior introduces more individualized treatment. An incentive individually optimized by a model introduces still more. And when that optimization is based on third-party information or an inferred willingness to pay, the distinction from the practices concerning the FTC becomes considerably harder to maintain.
The issue, then, isn’t personalization by itself. The area deserving particular attention is individualized economic treatment based on opaque personal-data decisioning that consumers would not reasonably expect.
The industry has an opportunity to define the line
The wrong response from the loyalty and incentives industry would be to insist that rewards aren’t prices and therefore none of this applies. The FTC is not currently proposing otherwise, but the economic boundary is not nearly as clean as the terminology suggests.
Nor is the answer to argue that personalized benefits deserve a blanket exemption because consumers receive something of value. As soon as individualized incentives create materially different effective economics — particularly when those differences are driven by information consumers don’t know is being used — that argument becomes difficult to sustain.
The better response is to help establish sensible distinctions while the FTC is still considering its policy. Direction of value matters. So does the basis for differentiation. Transparency and consumer expectations matter. And there are meaningful differences among a published loyalty benefit, a targeted promotion and an opaque individually optimized incentive.
Those considerations should also influence how companies govern their own systems. Companies should understand what determines who receives value and how much, how different the economic treatment of comparable customers can become, what data caused the difference and whether they can explain and reconstruct the decision. A useful test is whether the decision would still look reasonable if the consumer understood how it was made.
The FTC is accepting public comments on the proposed enforcement policy statement through September 18. That gives the loyalty and incentives industry an immediate opportunity to explain these distinctions while the Commission is considering how its personalized-pricing policy should work.
The useful argument is not “don’t regulate personalized incentives.” It is that individualized pricing and individualized benefits can produce very different outcomes for consumers, while acknowledging that the line between the two can become blurry. A workable framework needs to recognize both realities.
The FTC’s proposal is specifically about personalized pricing. But its treatment of individualized discounts, consumer expectations, disclosure and the use of personal data overlaps with practices already developing in loyalty and incentives.
That does not put the industry in the FTC’s crosshairs.
It does put the industry in the conversation.
And this is a good time to join it.
Photo by Michael Walter on Unsplash
