Retail consists of a few large blocks of hours that recur time and again: inventory management, pricing, customer contact on the shop floor and online, staff planning around peak moments, and the administrative handling of orders and returns. The circumstances that determine how those hours develop differ strongly per company: the ratio between physical stores and online sales, the extent to which assortment and prices are already adjusted automatically, and the question of whether customer contact mainly consists of repeat questions or of tailored work that requires knowledge of the store or the brand. Those differences also determine where AI is already taking over tasks today and where that is not yet happening.
In part of retail, AI is already taking over tasks today that previously fell to an employee as a matter of course: compiling order proposals based on sales data, answering frequently asked customer questions, flagging price deviations relative to competitors. In another part of the same sector, that is not yet happening, with the same technology available. The difference does not lie in the technology but in what a company considers reliable enough to let run without an intermediate step. Where an incorrect price adjustment or an incorrect order proposal is immediately noticeable in the till or on the shelf, human oversight remains necessary for longer; where the consequences of an error are small and quickly repaired, the work shifts more quickly to the category that runs without checks. Customer contact that revolves around exceptions, complaints or a store-specific arrangement remains, in most cases, human work - not because AI cannot handle the language, but because assessing what a customer needs in that situation does not fit into a fixed rule.
A strategy established two years ago was based on a certain distribution: so many hours of purchasing, so many hours of customer contact, so much FTE on the shop floor. That distribution was a reasonable assumption at the time. If AI has since taken over part of the purchasing preparation or the standard customer contact, that assumption has not turned out to be wrong, but has lapsed. No one has formally withdrawn the assumption; it still stands in the plan, while practice on the shop floor or in the warehouse has already moved on. That is the pattern by which strategy quietly ages: not through a wrong decision, but through an assumption that is no longer reconfirmed while the plan continues to rest on it.
The question is not whether AI is freeing up capacity in retail, because that is already happening, in parts and at varying speeds. The question for management is which assumptions in its own plan are built on that freed-up capacity, and whether those assumptions still match what is actually happening on the shop floor, in logistics and at the customer contact desk. That is a different matter from staff planning: what an employer does with freed-up hours or FTE capacity falls under its own statutory requirements for personnel decisions, and this page makes no statement about that. What matters here is the strategic layer above that: does the picture that management has of its own organisation still match the picture from outside.
That outside picture is not separate from what is happening in other sectors. Those who want to see how the same pattern occurs in hospitality will find that in how strategic pressure in hospitality is shifting due to AI; for the IT sector, where the shift often becomes visible earlier, the comparable analysis can be found in how AI is shifting strategic pressure in the IT sector. Those who want to understand what it means in practice to start looking from the outside world rather than from one's own plan will find that in what outside-in thinking means in practice.
The underlying question - which work in this company, in these stores, with this assortment, can actually be taken over by AI - is answered by the FTE TO AI work scan per task, rather than at sector level.
A strategy does not need to be revised all at once to be accurate again; it starts with seeing which assumptions underlie it and whether they still hold. Those who want to record why a particular strategic choice was made at the time will find an approach for that in how you record why you made a strategic choice. For those who want to start without immediately committing to a trajectory, there is the free assumption check: a short round in which you identify your most important assumptions and see, for each one, when it was last confirmed. The full pressure test, with the outside picture alongside management's self-plot, is under construction.
Stel uw vraag. Vaak zit de echte vraag een laag dieper — daar mag ik naar vragen.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.