The Price You See May Not Be the Price I See

Two people open the same shopping page and see different offers. It is tempting to conclude that one has been singled out to pay more. That conclusion may be right in a particular case, but the screenshots alone do not establish it. The price could have changed over time, a promotion could apply to one account, delivery could differ, or the page could be ranking different products rather than changing the price of the same one.

The more consequential possibility is personalisation: information about a person helps shape the price or promotion shown to that person. In January 2025, the US Federal Trade Commission published initial research on tools capable of using personal information in this way. That evidence deserves attention without becoming an accusation against every retailer or a claim that any two different offers prove unlawful conduct.

What the FTC actually studied

The FTC's 17 January 2025 announcement accompanied a staff perspective from its surveillance-pricing study. The work examined documents supplied by companies involved in pricing products and services. The agency described data including location, browsing behaviour and shopping history that could inform individualised prices or promotions.

The underlying staff perspective is important because it identifies what kind of evidence this was: initial findings from a study conducted under the FTC's information-gathering authority. It was not a final rule banning every form of personalised pricing, nor a finding that every named retail customer had used every capability discussed. Tools and commercial offerings are evidence of possible practices, not proof of an individual transaction.

The distinction prevents a familiar reporting shortcut. Saying that suppliers offer ways to use consumer data in pricing is not the same as proving that a particular shop raised a particular reader's price because of their inferred willingness to pay. That second claim needs transaction-specific evidence. This article makes no such allegation against a named retailer.

Four changes that can look like one

Dynamic pricing changes with circumstances such as time or demand. Personalised pricing depends on information about the person or audience receiving an offer. Promotions can be limited to an advertised membership or campaign. Ranking changes which product is displayed first. A single screen can contain several of those systems at once.

Consider a hypothetical pair of shoppers comparing headphones. One sees a lower headline amount but a higher delivery charge. Another sees a membership promotion with an eligibility condition. Neither example, by itself, proves a secret personal price. Now imagine the exact same product, seller, quantity, time and delivery terms producing different prices because one shopper's browsing history is used by the pricing system. That is a different proposition and a better description of the concern.

These are OLECH scenarios, not documented incidents. Their purpose is to show why the comparison needs controlled facts. A claim about the same product should identify the seller and product version. A claim about the same price should include mandatory charges and promotion conditions. A claim about personal data should establish what data affected the decision, rather than infer it solely from an account being logged in.

The data trail is the real story

The FTC material describes pricing intermediaries that can use detailed information about consumers and their behaviour. Even where the output is a promotion rather than a higher standard price, the commercial logic can be similar: determine which offer is most likely to produce a purchase. The relevant transparency question is whether people understand how their information enters that process.

A consumer may reasonably expect a store to remember an address or a basket. That does not automatically tell them that behavioural data could influence an offer. Nor does agreeing to a broad privacy notice necessarily make a particular pricing explanation easy to understand. This is an editorial observation about the gap between data collection and a person's understanding, not a legal conclusion about any individual notice.

Personalisation can also produce a lower offer. That possibility should not be edited out simply because a higher-price story is more dramatic. The issue is not that every personalised result is worse for every customer. It is that the basis for the difference may be hard to observe, compare or challenge. Benefits, disadvantages and opacity are separate dimensions.

A 2026 proposal is not a final rule

On 19 August 2026, the FTC issued a proposed enforcement policy statement on personalised pricing. The word proposed matters. Its text discusses a disclosure-focused enforcement position in circumstances where consumers reasonably expect prices not to vary based on personal information. It also states that the agency does not have authority to prohibit personalised pricing in all circumstances.

The FTC's 3 September extension announcement set the comment deadline at 25 September 2026. The policy-statement listing reviewed for this preparation continued to identify the item as proposed. A closed comment window does not transform a proposal into a final rule. No final adoption was established in the primary record reviewed.

Accordingly, this article does not describe the draft as an enacted ban, a binding new disclosure rule or a settled legal entitlement for every US shopper. It distinguishes the existing study, the proposed enforcement approach and any future final action. It also does not export a US agency's proposal to other jurisdictions as if consumer law were identical everywhere.

Why screenshots still have value

Careful comparisons can be useful even when they do not prove the whole mechanism. Record the product, seller, time, account conditions, delivery destination and final checkout total. If an offer is restricted, record the restriction. If the prices differ, ask what explains that difference before deciding that personal data caused it. Those are evidence-gathering principles, not a promise that an experiment reveals every hidden variable.

A different device or logged-out session may also change unrelated circumstances. Location, currency, inventory or an experiment run by the service could affect the result. The point is not to dismiss observed differences; it is to preserve the distinction between what was seen and why it happened. A confident causal story built from an uncontrolled comparison can obscure the very issue it is trying to expose.

For journalists and researchers, the stronger follow-up is documentary: the pricing policy, technical description, provider materials or a clear explanation from the seller. For shoppers, it is the complete payable offer and the conditions attached to it. The same evidence will not answer every question, but separating those questions makes the discussion less sensational and more useful.

A personalised internet needs legible prices

The internet already personalises feeds, recommendations and advertising. Extending that logic to prices raises a sharper issue because the output is not just attention; it is the amount a person may pay. The FTC's research shows why the capability deserves scrutiny. It does not establish that every retailer deploys it or that every personalised offer is harmful.

The honest conclusion is narrower than “everyone sees a secret price”. Some pricing systems can use personal information, and the US policy discussion includes a proposal about how that should be disclosed. Until the evidence identifies the practice in a particular transaction, different prices remain an observation rather than a complete explanation. The price you see may not be the price I see; the next question is what made it different.

Sources - FTC initial study announcement, 17 January 2025 - FTC staff perspective: research summaries - Proposed FTC enforcement policy, 19 August 2026 - FTC comment extension, 3 September 2026 - FTC policy-statement listing