Dollar General's Register Changed on Schedule. The Shelf Changed When Somebody Had Time. The Customer Was the Audit.

2026-10-29 · CX Pulse

The FTC says Dollar General tracked the price tags it printed, not the ones on the shelf. For three years the customer was the one checking.

On Monday the Federal Trade Commission and 20 states sued Dollar General in federal court in Tennessee. The short version is simple: for years, the complaint says, the register charged more than the shelf. The 95-page version describes how a price travels from headquarters to a receipt, and where along that route the company stopped looking.

Dollar General answered the same day. "Our goal is 100% accuracy between shelf and register prices in every store." It says discrepancies are occasional, that store teams are "instructed, empowered, and expected to make things right for the customer," and that it will defend the case. Its SEC filing says it has cooperated with FTC staff since February and disputes the agency's harm figures. Nothing in the complaint has been proven. The mechanics it describes are still the interesting part.

Two prices, two clocks

According to the complaint, Dollar General changes core prices weekly. The new price goes into the point-of-sale system and is charged at the register from then on, "regardless of whether shelf prices have yet been updated." The shelf is a separate job. Stores receive peel-and-stick labels on perforated sheets, sorted to the store layout, and are supposed to schedule a five-hour "Compliance Tuesday" shift to walk the aisles and hang them.

Core price changes are mostly increases. So whenever the register is ahead of the shelf, the customer pays more than the tag says. The complaint quotes a former assistant manager in Illinois, from an article that was sent to the CEO: "You're lucky if you get to work two to four hours of your eight- to 13-hour shift with another human being." A five-hour labeling shift in a store run by one person competes with the register, the truck and the floor.

Then the detail that carries the case. Headquarters keeps statistics on the labels that get printed. Per the complaint, those statistics "do not track whether the printed labels are actually affixed to shelves." The measurement stops one step short of the thing a customer reads. A former associate in Louisiana told a reporter she "threw away" new price tags "sometimes" to stay out of trouble, and nothing upstream would have shown it.

What the customer saw

The FTC's own test baskets were small. Four undercover buys of 58 to 60 items in Ohio, Missouri, Pennsylvania and Maryland. Error rates of 22.4%, 10.2%, 10% and 8.5%, against an inspection standard that fails a store above 2%. Net overcharges of $2.60 to $4.07 per basket, because some items rang up low. State inspections have the same shape: Kansas failed 39% of its checks, New Mexico 53%, and Iowa failed every one of eight.

A few dollars a basket, which is the point. The complaint describes checkout screens where prices "flash quickly," and one store where the running list faced the cashier and the customer saw only an abbreviated description. Some customers find the difference at home with the receipt. Most never find it. Between January 2023 and March 2026 the company logged 29,265 complaints in a category its agents call "Did Not Honor Marked Price," from every one of the 48 states it operates in. Those are the customers who caught it and took the time to say so.

Dollar General's policy is to honor the lower price, and two prior state settlements require it to post a sign saying so. When Wisconsin inspected 279 stores this winter, 35 had no sign. The complaint says that in many instances the cashier doesn't correct the price even when asked. "Make things right for the customer" is a real policy. It fires when the customer notices, asks, and the one person on shift agrees.

At scale: measure the copy the customer reads

Every large operation keeps a price, a policy or a promise in at least two places. The pricing engine and the shelf tag. The website and the store. The promo in the email and the promo in the cart. The copy the systems team owns gets updated on a schedule. The copy the customer reads gets updated when someone has time.

Three things to check this week. Where the accuracy number is measured: at the source, where the change was issued, or at the surface, where the customer sees it. "Labels printed" and "labels on the shelf" look like one metric until the second is sampled on the floor. The trigger for the make-good: a lower-price rule that waits for a raised hand pays out to the customers who notice; a register prompt when the scan differs from the shelf file, or an electronic label tied to the same price table, pays out to everyone. And the complaint category: if the help desk already has a bucket named for this, somebody built it because the volume earned a name. Read a month of it.

Smaller operation: the chalkboard is the shelf

The same gap lives in a six-person business, with the systems reversed. The price in Square, Shopify or the booking tool is live the moment it's saved. The menu board, the sign in the window, the price page and the quote in last month's email are the shelf, and nobody is scheduled to walk them. Change a price on a Tuesday, then walk the floor with a receipt on Wednesday. The customer who points at the old number shouldn't be the first time it comes up.

The complaint is an allegation and the court will decide it. The sentence that outlasts it is the company's own: the goal is 100% accuracy between the shelf and the register. The register has always matched the price file, which is what a register does. The shelf is where the customer stands, and for three years the customer was the one checking it.

Sources: FTC complaint, Case 3:26-cv-01693, M.D. Tenn., filed October 26, 2026 · FTC press release · Dollar General Form 8-K, October 26, 2026 · Dollar General's statement, via Fox Business