elmerdata.ai blog

My blog

The Price You See May Not Be the Price I See

What began as a nuisance in airline pricing is now raising deeper questions as the same logic approaches the grocery aisle.


From Airline Seats to Grocery Shelves

Airline passengers have long accepted a quiet truth: no two people on the same flight necessarily paid the same price. A seat on a plane became one of the first everyday examples of dynamic pricing, where timing, demand, and behavior shaped what each traveler paid. Experiences like those described in the JetBlue post have become routine, even if they still frustrate. That model is now moving beyond travel. See previous blog entry.

In Maryland, lawmakers have advanced legislation aimed at restricting personalized pricing in grocery stores, a practice often described as surveillance pricing. Surveillance pricing is the practice of setting or adjusting the price of a good or service for an individual consumer based on personal data collected about that person, including browsing history, past purchases, location, and inferred willingness to pay. These systems may rely on personally identifiable information, such as loyalty accounts, or operate through behavioral profiles that effectively function the same way. Retailers increasingly have the ability to adjust prices based not only on demand, but on who the shopper is, what they have searched for, where they live, and how likely they are to complete a purchase.

Such capabilities reflect a deeper shift. Pricing is no longer just a function of supply and demand. It has become a function of data, prediction, and individual profiling. Historically, grocery stores operated on a shared understanding: the price on the shelf applied to everyone. Discounts existed, but they followed visible rules. That consistency built trust over time. The emerging model challenges that tradition. Two shoppers standing side by side could encounter different prices for the same item, not because of a posted promotion, but because an algorithm has inferred their willingness to pay. The logic that shaped airline tickets is now approaching essential goods.

groceries Dean Hochman, Hy-Vee Grocery Interior, Overland Park, Kansas, 2013. Licensed under CC BY 2.0.


A Law That Signals Concern, Not Resolution

The debate unfolding in Maryland is not simply about technology. The tools already exist. The question is whether they belong in places where expectations have been stable for generations. The proposed law, known as the Protection From Predatory Pricing Act, attempts to prohibit grocery retailers and food delivery services from using personal data to set individualized prices. On its face, it reflects a traditional principle: the price on the shelf should be the price paid at the register.

Yet the details suggest a more complicated reality. Analysts at Consumer Reports and policy experts warn that the bill contains significant limitations. It applies only to food retail, leaving other sectors, including airlines and apparel, untouched. It also includes broad exemptions for loyalty programs, subscriptions, and cases where consumers consent to data use in exchange for pricing changes. Enforcement raises further questions. The law relies solely on the state attorney general to bring cases and allows a 45-day window for violations to be corrected, creating uncertainty about how strongly it can be applied in practice.

Taken together, these provisions suggest that the legislation is less a definitive solution and more an early attempt to draw boundaries around a rapidly expanding practice. Even so, the signal is clear. Dynamic pricing, once confined to airlines and hotels, is pressing into domains where consistency has long been the norm. Lawmakers are beginning to respond, even if imperfectly, to a shift that challenges long-standing expectations of fairness and transparency.

The question now is not whether personalized pricing will spread. It already has. The question is whether institutions can preserve a shared understanding of price in the face of technologies designed to make pricing increasingly individual. That tension, between efficiency and fairness, will define the next phase of this debate.


Commentary: Why does this feel wrong if Amazon already does it?

The obvious objection is simple. If pricing already shifts on platforms like Amazon, why draw a line now? Consumers have lived with changing prices for years. Airline tickets rise and fall. Online listings update constantly. Nothing about that feels new.

The difference is not that prices move. The difference is how they move. Most dynamic pricing still operates in public view. Prices change over time, but everyone sees the same price at a given moment. That shared reference point preserves a basic sense of fairness.

Surveillance pricing breaks that norm. It introduces the possibility that the price is not just changing, but changing for you. A shared market price becomes a private calculation. That shift is subtle, but it alters the relationship between buyer and seller in a fundamental way.

Groceries make the issue harder to ignore. Food is not discretionary. It is routine and necessary. A system that quietly charges different people different amounts for the same staple item conflicts with a long-standing expectation that essential goods are priced in a way that is consistent and knowable.

There is also a practical concern. When prices are individualized and opaque, comparison shopping loses meaning. Each consumer operates in a slightly different reality, shaped by data they cannot see and decisions they cannot challenge.

Platforms like Amazon may have normalized price fluctuation, but they have not fully normalized individualized pricing for essential goods. The response in Maryland suggests that boundary still matters. The discomfort is not nostalgia. It reflects a simple principle: markets function best when prices are visible, comparable, and shared.


Further Reading

NPR story -->

Maryland's Bill -->


AI Assistance Statement ▾
Preparation of this blog entry included drafting assistance from ChatGPT using a GPT-5 series reasoning model. The tool was used to help organize ideas, propose structure, refine language, and accelerate revision. It was also used to assist in identifying image sources and verifying that selected images appear to be released for reuse (for example through public domain or Creative Commons licensing). The author selected the topic, determined the argument, reviewed and edited the text, confirmed image licensing, and takes full responsibility for the final published content. (Last updated: 03/06/2026)

#AIData #Observations