LO
Retail cashier / shop assistant — tasks, one by one
The unit of analysis is the task, not the job title. Each one below carries its direction, whether the judgement rests on evidence or on platform inference, the reasoning, and what it does not establish.
Every task on this page#
Scanning and taking payment
Automating✓ Evidence-backedThe till: scanning, bagging, payment, receipts.
Self-checkout, scan-as-you-shop and checkout-free formats have been deployed at scale for over a decade and have reduced staffed tills in most large chains. The technology is mature and the retailers' motive — labour is their largest controllable cost — is not going away.
Several major retailers have pulled back self-checkout after theft and customer frustration outweighed the savings. Adoption has oscillated, which is a useful reminder that deployment is a business decision, not a technology inevitability.
Restocking and shelf work
Still human-led≈ Platform inferenceUnpacking deliveries, filling shelves, facing up, rotating dates, handling the 3,000 products that are each slightly different.
Shelf-scanning robots exist and count stock; robots that restock a mixed shelf in a live store do not exist at commercial scale, because the objects are varied, fragile and packed by someone else, and the store is full of people. This is the largest block of physical work in the role and it has proved the hardest to automate.
Shelf work is physically hard, low-paid and scheduled around peaks. Being the least automatable task in the shop is not the same as being a job people can build a life on.
Helping customers
Still human-led≈ Platform inferenceWhere is it, is there more in the back, which one is better, can I return this, my card did not work.
In-store help is physical and immediate, and it is the reason a physical shop beats a website for many purchases. Retailers that cut floor staff to fund self-checkout have measured the effect on satisfaction and sales, and several have reversed course; the person on the floor is part of the product.
Retailers measuring the effect on satisfaction and reversing course is one set of retailers. Others measured the same thing and kept the savings, and the site has evidence of both.
Loss prevention and the difficult customer
Still human-led✓ Evidence-backedNoticing theft, handling the argument at self-checkout, de-escalating, deciding when to call for help.
Self-checkout shifted work from scanning to supervising, and the supervising is harder: it is about judgement, confrontation and safety. Cameras and analytics flag; a person decides what to do about it, in a situation that can become dangerous.
This task grew because self-checkout created the theft it now polices. It is work generated by an automation decision, and it is more dangerous and no better paid than the work it replaced.
Picking online orders in store
New task≈ Platform inferenceWalking the aisles to assemble a customer's online basket for collection or delivery.
Online grocery and click-and-collect turned stores into fulfilment centres, and picking is now a large share of hours in many supermarkets. It is new work created by the same digital shift that reduced tills, and it is physical work in a live store that automation has not reached.
In-store picking exists because delivery economics have not been solved, not because the work is valuable. Dark stores and automated fulfilment centres are the direction of travel where volume justifies them.