As of 9 November 2026, the most consequential pricing document in American retail is a set of questions rather than a rule. On 28 October the Federal Trade Commission voted 2-0 to issue Section 6(b) orders to nine companies, including Walmart, Target, Instacart, DoorDash, Uber, Lyft, Delta, United and JetBlue, asking how personal data is used to set individual prices and, on the gig platforms, individual pay. The commission has made no finding of wrongdoing, and the practical effect on a category manager is simpler than the headlines suggest: price monitoring now has to be documented rather than assumed.
What the orders actually require
Read the scope carefully and the shape of the request appears. Companies must describe each methodology that uses personal information to set or adjust a price, name the algorithms and automated decision tools involved, explain how inputs affect the outcome, and identify which categories of consumer are treated differently. That is a data-governance question dressed as a pricing question, and most retail organisations cannot answer it today from a single system.
The timing matters too. The orders land on top of an August 2026 proposed enforcement policy statement on personalised pricing, whose comment period was extended to 21 November, and alongside pricing bills that roughly 30 state legislatures have considered this year. The direction is consistent across federal and state activity, which is why a retailer should treat the 45-day clock as the first of several deadlines rather than a one-off.
Why a study changes price monitoring
Price monitoring used to mean checking whether a reseller undercut a recommended price. The question now includes whether the same item shows a different price to two logged-in shoppers, and that cannot be answered by scraping a product page once a day. It requires comparing the app while logged in, the site while logged out and the shelf label in a store, then keeping the evidence with a timestamp.
The three views that have to match
Each view fails for its own reason. App prices can carry a loyalty or membership adjustment, logged-out web prices are the closest thing to a public reference, and shelf prices carry local and promotional differences that a national feed will never show. A price gap between two of the three is not proof of anything, but it is a question with a documented answer.
| Channel compared | What a difference usually means | Evidence to keep |
|---|---|---|
| App logged in versus web logged out | Membership or loyalty adjustment, or a test | Screenshot with time, account tier and device |
| Web logged out versus store shelf | Regional pricing or an unrefreshed shelf label | Photo of label plus store code |
| Marketplace listing versus brand store | Reseller pricing or a different pack | Listing URL, seller name, pack size |
| Promotional price versus 30-day median | Discount depth, real or inflated | Daily price series for the item |
The fourth row is where most disputes actually start, because a personalised price and a legitimate promotion look identical from outside. The only way to separate them is a reference window long enough to show what the item normally costs, which is why the 30-day median keeps reappearing in pricing rules across markets.
What a retailer should document before the deadline
Walmart's response sets the tone for the industry: in a 5 October letter its executive Dan Bartlett wrote that the company prices products, not people and does not use AI to upcharge customers. Instacart says prices on its platform are set by the retailers it serves, and that shopper pay depends on distance, items and expected shopping time. Both statements are defensible only if the underlying records exist.
Disclosure is a data question
If a retailer cannot list the inputs that feed a price, it cannot claim the list is limited to product and cost. The disclosure package the orders require, covering the language used, where it appeared and when, is effectively an audit of the pricing stack. Building it once is also the cheapest way to answer a state inquiry or a class action later.
How to run a price audit in four weeks
Four weeks is enough to produce evidence, not a legal opinion. The audit works only if the same items are read at the same hour every day, because a pricing test that runs on weekday mornings will never appear in data collected at random. Assign one owner per step, keep the raw screenshots rather than a summary, and treat every gap as a question to answer before it becomes an explanation. The detailed sequence and the record format sit below, hidden from the page layout so the reading stays on the argument.
FAQ
Does the inquiry mean personalised pricing is illegal?
No. The orders are a fact-finding study under Section 6(b), and the commission has made no finding against any company, so the immediate risk is disclosure and record-keeping rather than penalty.
Why does worker pay appear in a pricing inquiry?
Because gig platforms set both the consumer price and the offer shown to a driver, and the commission wants to know whether the same personal data influences compensation for comparable work.
Which check can a brand run without regulator access?
Comparing the same item in the app, on the logged-out website and on the shelf, on the same day, and keeping the screenshots, which is enough to raise a specific question with a retailer.
What happens after the 45 days?
Commission studies typically produce a public report after several months, and reports have historically been followed by policy statements, further inquiries or litigation.
