Walmart pledges no personalized pricing. Write yours down.

Walmart just made its pricing algorithm a public promise. On September 25, CEO John Furner published a letter committing the company to never set a price from who you are — not income, not shopping history, not urgency — and extending that rule to its Sparky AI assistant and its electronic shelf labels. The line to remember: "We price the product, not the person." If you sell anything online, personalized pricing is now a thing you have to have a position on, in writing, whether or not you have ever done it.

What actually happened

Furner's letter makes three commitments. Prices do not change based on identity or time of day. Information a customer shares with Sparky will not be used to raise their price or bury a cheaper option that would have worked. And the digital shelf labels — the thing that set off the rumor cycle — exist to keep the shelf price matched to the register price and to stop associates from swapping paper tags by hand. Walmart says those labels carry no cameras, no microphones, no facial recognition.

Nothing in the letter is technically new. What is new is that a retailer of that size decided a denial was cheaper than the ambiguity.

The timing is not an accident. On August 19 the FTC proposed an enforcement policy statement on personalized pricing. Per DLA Piper's read, the Commission concedes it cannot ban the practice outright, but takes the position that a business which does not tell customers how their personal data feeds a price may be violating the FTC Act. That reframes the whole problem. The exposure is not the algorithm. The exposure is the silence around it.

States are moving on a separate track. Greenberg Traurig's survey describes Illinois HB 4248 as a disclosure-plus-opt-out model: tell customers which categories of personal data go into a price, and let them decline and still get the base price. Outright bans have been introduced and have mostly failed. Disclosure rules are the ones that pass.

Why personalized pricing matters for your business

Here is the uncomfortable part. You may already be doing this without calling it that.

A logged-in cart that surfaces a different price than an anonymous one. A retargeting flow that emails a 15% code to people who bounced twice. A Shopify Function that reads customer tags and swaps the tier. A dynamic pricing app you installed in 2023 and never audited. Every one of those is a price that moved because of something you know about a person. Some of it is a loyalty discount and perfectly defensible. Some of it is surveillance pricing with better branding. The FTC's proposed standard does not care which word you use — it cares whether the customer was told.

So do the boring work now, before someone asks you in a demand letter.

Inventory the inputs. List every system that can change a displayed price: apps, discount rules, tags, segments, A/B tests, personalization scripts. For each one, write down which customer attribute it reads. If nobody on your team can produce that list in an afternoon, that is the finding.

Publish your version of the Furner letter. One page, plain language, on your own domain. What you use data for, what you never use it for, how a customer opts out. Walmart's move works because it is specific and findable. A vague privacy policy paragraph is not the same artifact.

Separate discounts from prices. A code anyone can use, a published loyalty tier, a volume break in your quote table — those are discounts, and they survive scrutiny because the rule is visible. A silently different base price does not. Restructure toward the first category.

Sources: Walmart CEO letter, September 25, 2026, DLA Piper on the FTC's proposed personalized pricing policy statement, Greenberg Traurig on state restrictions.