Retail salesperson / shop assistant

Total retail trade, all employees, seasonally adjusted. The long view is the finding rather than the recent months: this sector has employed roughly the same number of people for seven years, through the largest shift to online ordering in its history. That is a fact about retail employment and not about this occupation specifically — the series contains cashiers, stockroom staff, managers and delivery, and cannot separate shop-floor selling from any of them. It also cannot see the warehouse and logistics jobs that e-commerce moved into other industries, so it is not a measure of where retail work went.

Verifiable change in hiring, headcount, hours or job scope. Highest weight — but causal attribution still has to be argued, not assumed.

The company's own announcement. It says it had not yet created a truly distinctive customer experience with the right economic model for large-scale expansion and would close its Amazon Go and Amazon Fresh physical stores, while its Just Walk Out checkout-free technology operates in over 360 third-party locations across five countries. The company is both the store operator and the seller of the checkout-free system, so the two halves carry different stakes.

An employer has put it into production. Can move the baseline — weighted by scale and how similar the setting is.

The retailer's own account of its checkout changes. It says that in March 2024 it rolled out express self-checkout lanes limited to 10 items or fewer at most of its nearly 2,000 US stores, opened more traditional lanes staffed by its team for full carts — more than 25,600 nationwide — and reports transaction times improving by nearly 8%. It is the company's own promotional account and gives no staffing change.

An employer has put it into production. Can move the baseline — weighted by scale and how similar the setting is.

The retailer's announcement of an AI assistant built for its associates' sales-floor devices. It says the app provides fast access to product details, project advice and inventory information, gives an example of an associate asking which fertiliser works best for Bermuda grass, and says it launched to all associates across more than 1,700 stores. It is the company's own release, with its AI developer quoted as a partner, and it gives no usage figures or effect on sales staff.

An employer has put it into production. Can move the baseline — weighted by scale and how similar the setting is.

California. The bill would have prohibited grocery stores and retail drug establishments from offering self-service checkout unless conditions were met, including having no more than two self-service checkout stations monitored by any one employee and requiring that employee to be relieved of all other duties, and would have required notice before introducing consequential workplace technology. It passed the Senate in May 2024 and its last recorded action is 'From Assembly without further action' on 30 November 2024. This record establishes that the rule does not exist, not that California has no other rules on checkout.

A rule that would have constrained or required automation was formally proposed and did not come into force — vetoed, voted down, struck down, or allowed to lapse. It establishes something real and checkable about the legal environment: a guardrail many people assume exists does not. It never moves a task's assessment, because it says nothing about what employers can do or are doing.

The retailer's annual report to US securities regulators. It says it is revising its self-checkout strategy, including limiting self-checkout to transactions of five items or fewer, converting some or all self-checkout registers in approximately 9,000 stores to assisted checkout options and removing self-checkout from approximately 300 stores, alongside investment in retail labour; it names self-checkout among the operational challenges behind higher shrink in 2023. It is one company's decision about its own stores and says nothing about how many staff the change added.

An employer has put it into production. Can move the baseline — weighted by scale and how similar the setting is.

This is still one of the easiest jobs in any economy to get, and that has not changed. What has changed is where it leads: the stock-knowledge rung that used to make someone indispensable in their second year has been flattened by a customer-facing inventory app. Aim at the categories where the product has to be tried, fitted or explained, because those are the ones where the floor keeps its people.

Your leverage is the conversation, and the risk is that it gets relocated rather than replaced — into a chat channel where one person covers eight stores. If your employer is building that channel, the question to ask is who staffs it, because that is the same job with a different postcode and usually a different pay band.

Reshape the role

Move to a category that must be tried on

Where the product has to be fitted, demonstrated or sized, the sale cannot happen without proximity, and proximity is what this occupation has.

Adjacent move

Follow the stock, not the till

Stores that have become collection points for online orders need someone who owns accuracy between the app and the shelf, and that person is paid more than the floor.

Not in the way the question implies, and the more useful split is between the two halves of the job. Answering what is in stock has already largely gone, because the customer can look it up — that was the errand that started most conversations. Working out what someone actually wants has not, because it depends on watching them handle the thing, which no recommender receives as an input. The realistic risk is not a robot on the floor; it is the conversation being relocated to a chat channel where one person covers several stores.

No date. The signal for this job is whether your store still holds stock a customer can touch, because everything durable about the role is anchored to that. Watch the floor space given to collection lockers versus display, and watch whether your shifts are being set to the half hour against a footfall forecast — that can cut hours without removing a single task, and it can reach you before any robot does.

Nothing on this page records a shelf-scanning robot, so read this as reasoning: a robot can audit a shelf but cannot fix it, so a person still walks the same aisle with a trolley. What a scanning robot changes is who knows the shelf is wrong, not who corrects it — and correcting it is where the hours are.

No, and treating them as one is how people get the wrong answer about both. Cashiering is a transaction task with the longest-running automation in retail pointed straight at it. Selling on a floor is an advisory task anchored to physical stock. Many people do both in the same shift, which is exactly why the exposure of the shift is not the exposure of either task — and why this site keeps them on separate pages.