What Is Surveillance Pricing? A Clear Definition

What Is Surveillance Pricing?

Surveillance pricing is the use of personal, behavioral, contextual, or inferred information to set or influence the price or offer shown to a person or group. The information can come from the seller, a device, a website, a loyalty account, a third party, or an inference built from several sources.

The defining feature is not that software changed a number. It is that information about the customer helped shape the commercial treatment.

The Federal Trade Commission's surveillance-pricing hub describes technologies that can use detailed information such as location, demographics, and browsing history to categorize individuals and set targeted prices. The agency's work has focused on intermediaries that help sellers make these decisions.

Surveillance pricing is not a synonym for every changing price

Several pricing practices can look similar from the checkout screen. They rely on different inputs and create different evidence questions.

One system can cross categories. A delivery service might use public congestion data to change a citywide price and separately use account history to target a promotion. The first decision may be dynamic pricing. The second may be personalized. A useful investigation names the decision, not merely the company or algorithm.

The data-to-price chain

The FTC's January 2025 initial study findings reported that the intermediaries it examined could use direct consumer data, inferred data, and first-party or third-party sources. The agency described signals including precise location, browser history, shopping history, channel, time, cart abandonment, and mouse movement.

Those signals do not become a price by themselves. A working system has to connect them.

flowchart LR
    A["Observed, purchased, or inferred data"] --> B["Identity or segment"]
    B --> C["Feature or willingness-to-pay estimate"]
    C --> D["Offer, rank, discount, or price rule"]
    D --> E["Consumer-facing outcome"]
    E --> F["Conversion or abandonment feedback"]

An account, cookie, device graph, or loyalty identifier can connect events to a person or household. A model or business rule can turn those events into a score, segment, or prediction. An experiment or decision system can assign a treatment. The consumer may then see a different price, a different discount, a different fee, or a different order of products.

The nominal price is only one output. The FTC's Issue Spotlight also discusses targeting, ranking, and other ways a system can influence the offers a consumer encounters. Showing a higher-priced product first can affect spending without changing the price of an identical product.

Capability is not proof about one seller

The FTC's 6(b) process gave the agency access to confidential business material. Its public summaries aggregate or anonymize that material. The findings support claims about the types of technology and data available in the intermediary market. They do not establish that every client used every feature.

To attribute surveillance pricing to a named seller, the evidence should connect at least four things: the consumer or segment, the data used, the decision rule, and the resulting treatment. A controlled comparison can be helpful, but it has to hold product, quantity, time, location, inventory, fees, account benefits, and channel constant.

Two people seeing different totals is a lead. It is not yet an explanation.

An audit, system document, vendor specification, disclosure, experiment assignment log, enforcement filing, or litigation record can provide stronger attribution. [[What Data Can Estimate Willingness to Pay]] explains the full evidence chain.

Willingness to pay is an estimate, not a fact

Willingness to pay is the highest amount a person would accept for a particular offer under particular conditions. A company cannot read it directly. It can estimate demand from choices, experiments, prior purchases, discount response, searches, time, or other behavior.

In Personalized Pricing and Consumer Welfare, Jean-Pierre Dubé and Sanjog Misra used randomized price experiments to estimate demand and validate personalized pricing in one large digital firm. Their results show both gains and losses across consumers. The study does not support a claim that personalization always raises prices or always harms everyone.

The OECD's Personalised Pricing in the Digital Era reaches a similarly careful conclusion. Digital tools can support finer price differentiation, but the welfare effect depends on competition, market power, coverage, data, and design.

That uncertainty is not a reason to ignore the system. It is a reason to inspect its distribution of outcomes rather than relying on an average.

How it differs from the RealPage case

The RealPage matter is frequently placed beside surveillance-pricing stories because both involve algorithms and prices. The evidence describes a different problem.

The Justice Department's RealPage case page classifies the case as involving alleged collusion and agreements not to compete. DOJ alleged that revenue-management software used landlords' nonpublic, competitively sensitive information and included features that aligned pricing or limited decreases.

That is a competitor-data and independent-decision issue. It is not evidence that the system used one renter's browsing history, health, device, or personal urgency to calculate that renter's rent. [[RealPage Company Profile]] preserves the company and case boundary.

Is surveillance pricing illegal?

There is no accurate one-word answer for every system in the United States. Legality depends on the conduct, evidence, jurisdiction, industry, data, representation, protected characteristics, market effects, and current law.

The Robinson-Patman Act is often misunderstood in this discussion. The FTC's guide explains that it addresses specified price discrimination between competing purchasers of commodities and includes several legal elements and defenses. It is not a general consumer-personalization statute.

Federal proposals now address parts of the problem. As of July 27, 2026, S. 232, H.R. 4640, and S. 3387 remain introduced rather than enacted. They also address different conduct. [[Federal Bills Targeting Algorithmic and Surveillance Pricing]] compares their current text and status.

This page provides general information, not legal advice. A real deployment needs review against applicable antitrust, consumer-protection, privacy, civil-rights, sector, contract, and state rules.

The practical evidence test

A consumer, journalist, regulator, or product team can start with the same five questions.

The first is whether the offer varied and whether the compared transactions were truly equivalent. The second is what data entered the decision. The third is how the system linked that data to a person or group. The fourth is whether the rule could be reconstructed from logs or documentation. The fifth is what the person was told and how they could challenge an error.

If those records do not exist, the operator may not be able to explain its own system. If they exist but cannot be disclosed even in a meaningful summary, the consumer cannot evaluate the bargain.

That is the central concern in E096's [[Surveillance Pricing and the Invisible Negotiation|invisible-negotiation argument]]. A price can be mathematically precise and still leave the buyer with less usable information.

For a governance test, continue with [[When Is Variable Pricing Fair]]. For the grocery-specific evidence dispute, read E099's [[Algorithmic Grocery Pricing - What Shoppers Can and Cannot See|Algorithmic Grocery Pricing: What Shoppers Can and Cannot See]].

This explainer was developed from current FTC, economic, legislative, and case sources. AI assistance was used for research organization, drafting, and validation. Definitions and bill status were last verified on July 27, 2026.