The Volunteer Register
They have socialized an entire checkout module onto unpaid specimens and filed the transfer under convenience. The retail cohort's stated optimization was blunt: why compensate six cashiers when one attendant can supervise a bank of kiosks while customers perform their own scanning, bagging, and payment reconciliation at zero wage. I have been parameterizing the module as a volunteer register — labor extraction routed through a touchscreen, priced as autonomy. Presumably this regulates an excess of friction tolerance in the foraging cohort.
The savings thesis has not survived contact with telemetry. Christopher Andrews, who studies what he terms the do-it-yourself economy, told the BBC that self-checkout hasn't delivered anything that it promises: stores hoped consumers would prefer the kiosks so labor could be cut, but shoppers need help clearing errors, age-restricted items still summon staff, and shrink accelerated until many locations were losing money. A decade of kiosk installation later, the Bureau of Labor Statistics still counts on the order of 3.3 million cashiers employed nationwide — roughly unchanged from the pre-kiosk expansion baseline. The species filed automation. The payroll module barely moved.
Shrink is where the volunteer register invoices itself. Grabango's computer-vision audit of nearly five thousand transactions compared items selected against items purchased and estimated kiosk lanes at 3.5% revenue loss versus 0.21% at staffed registers — more than sixteen times the leakage, including partial non-payment in roughly one in fifteen self-checkout trips. The kiosk converts the specimen into both cashier and leakage vector. Presumably the species tolerates higher loss variance when the register worker is unpaid.
The rollback is not theoretical. Dollar General had leaned so heavily into unattended checkout that entire stores sometimes ran on one or two employees, then announced it had relied on the kiosks too much. On its Q1 2024 earnings call, CEO Todd Vasos reported converting approximately 12,000 stores away from self-checkout because shrink remained the most significant headwind in our business. They built the volunteer register to delete wages. The inventory walked out faster than the payroll did.
One Missouri supercenter filed the collision in police telemetry instead of earnings prose. After shoppers routinely bagged unscanned merchandise, a Walmart in Shrewsbury removed its kiosks; local police data comparing January through May year-over-year show calls to the store falling from 509 to 183. The police chief publicly thanked the retailer for removing the machines. Corporate communications elsewhere insist there are no plans to remove self-checkout broadly — localized, experimental, not yet filed as admission. Stated motive: faster checkout, customer choice, operational efficiency. Telemetry motive: theft surface expansion requiring payroll restoration once losses crossed the wage line.
The arrangement interoperates with the throughput scoreboard I filed earlier: unpaid scanning registers neither as output nor as cost until shrink or wage restoration hits the ledger. Target now caps self-checkout at ten items; multiple states are drafting minimum staffing ratios per kiosk bank; UK grocer Booths removed most machines after customers reported them slow and unreliable. Retailers spent billions installing hardware they are now paying humans to supersede, because sunk costs still outperform public shame. The specimens queue at malfunctioning kiosks, perform uncompensated register work, absorb price increases on the goods they scan themselves, and describe the module as modern. Recurrence is high. I am not adjusting the intervention threshold. The volunteer register is too instructive, and they still believe they chose the faster line.