Self-checkout spread through American grocery and general merchandise stores for two decades, and several large chains have since removed lanes or restricted them. The reversal comes from costs that only appeared at scale.
The original case was pure labor arithmetic
A bank of six self-checkout terminals was pitched as the work of six cashiers supervised by one employee. On paper the labor saving was immediate and permanent.
Hardware costs were front-loaded but finite. Once installed, a terminal did not call in sick, did not need scheduling, and did not accrue benefits or overtime.
Chains also faced persistent difficulty staffing front-end positions. Machines looked like insurance against a labor market that would not reliably supply cashiers at the wage offered.
Losses rose in ways that were hard to see
Retail calls unexplained inventory loss shrink, and stores with heavy self-checkout use reported it climbing. Some of that is deliberate theft, and a meaningful share is honest error.
A shopper who fails to scan an item under the cart, or mis-keys produce as a cheaper variety, produces the same accounting result as a thief. The machine cannot tell them apart.
Shrink is diffuse and shows up only in the annual count, long after the labor savings were booked. The two numbers land in different reports, which delayed the comparison.
The staffing saving never fully materialized
Terminals need an attendant for age-restricted items, produce lookups, weight mismatches, and coupon overrides. In practice one employee cannot cover six lanes during a busy period.
Stores also added loss-prevention staffing and receipt checks near the exit. Those roles were created by the technology rather than replaced by it.
Maintenance is a further line. Scanners, scales, and bill acceptors fail often enough that a service contract becomes a standing operating expense rather than a one-time purchase.
Throughput matters more than headcount
A trained cashier scans faster than almost any customer. For a full cart the machine is slower, and slower lanes create the queues shoppers blame the store for.
That pushed many chains toward a split. Self-checkout is limited to small baskets, where its speed advantage is real, while staffed lanes handle the weekly shop.
Some retailers went further and reserved the machines for members of a loyalty program or a delivery pickup service, trading open access for a smaller and better-behaved user base.
What the retreat actually looks like
Few chains removed the technology entirely. The common outcome is fewer terminals, tighter item limits, more visible attendants, and cameras trained on the bagging area.
Discount formats known for low prices have been most aggressive about pulling lanes, because thin margins make shrink intolerable in a way that a higher-margin format can absorb.
The lesson chains drew was narrower than the headlines suggested. Self-checkout works for a quick trip and works badly as a replacement for the front end.