Retail is a sector with two faces. On the shop floor, the work revolves around physical actions: stocking shelves, moving products, helping customers, staffing tills, keeping spaces clean. At the office or head office, the work is quite different: purchasing, inventory management, pricing, marketing, staff planning and administration. In most companies these two worlds run through each other, which makes a generic statement about "AI in retail" not very useful. The outcome depends heavily on which part of the work you look at.
In this type of business, most office hours go into recurring, information-driven work: matching orders to sales figures, adjusting prices, compiling reports, communication with suppliers and customer service via email or chat. On the shop floor, most hours go into physical work and direct customer contact, which a screen or algorithm can take over very little of.
Three circumstances strongly determine what AI can and cannot do here:
Within retail, there are tasks that lend themselves well to full automation, especially when they revolve around fixed rules and repetition. Think of generating standard reports on sales and inventory, answering frequently asked customer questions via chat, drafting initial versions of product descriptions, and flagging price discrepancies between channels. Simple order suggestions based on historical sales data often also fall into this category, provided the underlying data is reliable.
A large part of the work in retail falls into the intermediate category: AI makes a proposal, an employee approves or rejects it and gives a reason. Examples include dynamic price adjustments, where an algorithm proposes a price but a buyer must assess the context of a promotion or competitor; inventory replenishment amid fluctuating demand, where an employee factors in seasonal influences or local events; and customer contact that goes beyond a standard question, such as a complaint about a defective product. This is also the category where most discussion arises, because AI's proposal is often good, but not automatically correct for the specific situation.
Some tasks remain human work, not because AI cannot handle them, but because the nature of the work calls for it. Physical actions on the shop floor, building a relationship with regular customers, managing a team on the floor, and handling complex or emotionally charged complaints usually fall into this category. Decisions carrying legal weight, such as assessing a liability issue or a labour dispute, also belong here: these are subject to their own statutory requirements and fall outside the scope of an AI assessment.
To determine what shifts in a specific company, it helps to look beyond the shop floor alone. Supply and transport to branches, for example, touch on tasks also described in what work can AI take over in logistics, while staff planning for varying shifts and peak moments overlaps with what is described in what work can AI take over in planning. Companies with their own distribution centre or wholesale position may also recognise themselves in the tasks described in what can AI take over in wholesale. Those who also want to include personnel files, absenteeism or recruitment in the overview will find additional context in what work can AI take over in the HR department.
This classification is a description of tasks, not a justification for personnel decisions. Whether and how an organisation does something with freed-up hours is up to the organisation itself. Decisions affecting positions or contracts are subject to their own statutory requirements, which this page does not provide for.
The classification into these three categories is a starting point, not a final verdict: the precise distribution depends on the systems, the customer group and the way of working in a specific company. If you would like an initial indication for your own situation, you can fill in ftetoai's free quickscan: twelve questions, no account needed, with an indication of what portion of the hours in your profile can be taken over by AI today. The full work scan, which goes deeper into individual tasks and processes, is still under construction.
Vraag maar. Ik ken de kennisbank van deze site; wat ik niet weet, zeg ik erbij.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.