Currys: selling the washing machine is easier than getting it working

A washing machine can be perfect on price, colour and number of programmes. Then the doorway is too narrow, the space is wrong or the connection requires work nobody discussed. Specifications provide limited comfort at that point. The customer bought the ability to wash clothes and received a large object to negotiate around.

Currys offers home installation services. That is the documented starting point for my independent AI sales concept. I am not describing a Currys deployment, nor assuming that every service applies to every product and address. The proposal follows the purchase through to actual use.

Sell a workable outcome

Before payment, an assistant checks the available space, access route, connection type and whether an old appliance needs removing. Each question explains why it matters. If the customer does not know, the workflow requests clarification or an appropriate assessment. It must not fill missing fields with optimism.

AI compares those answers with the specific model's instructions and the selected service conditions. The result separates confirmed requirements, unresolved checks and excluded work. Customers see the boundaries of the installation, not just the appliance price. That is more useful than discovering the relevant small print beside a delivery box.

If additional work may be necessary, the associate offers an available assessment route. They should not automatically add it to the basket or promise that the installer will sort everything out on arrival. Customers usually finance that sort of optimism with their own time. Clear agreement before payment protects the relationship.

AI is not a remote installer

The reference library needs current dimensions, manuals, requirements and local service terms. Technical or safety uncertainty goes to qualified people. The assistant must not instruct customers to modify gas or electrical systems themselves. Initial screening does not require photographs of the entire home or unrelated personal information.

Test one appliance category in a limited service area. Installation colleagues help design the questions around actual causes of failed visits. Each failure is reviewed to distinguish a missing question from an unclear answer or a later change in conditions. Those are different problems, not one generic prompt failure.

The installation operation provides the evidence: installation completed on the agreed first visit, cancellations caused by undiscovered requirements, and repeat contacts about service scope. Additional-service revenue should not be the primary success criterion. The purpose is fewer surprises, not more opportunities to make the opening price less representative of the final bill.

An appliance sale is not really complete when the receipt prints. It is complete when the person can use what they bought. That is the unit of service I would design around. AI gathers relevant questions, specialists own technical decisions, and the customer receives clarity. It sounds unremarkable until you calculate the cost of its absence.