A customer who reports a fault, reorders a part or asks about maintenance is not, at that moment, comparing your product to a competitor's. He is comparing the experience of being helped. How quickly someone responds, how well that person knows the installation, whether the problem is solved in one go. For many customers, that experience has come to weigh more heavily than the purchase itself, because the purchase is a one-off and the service spans the entire lifetime of the product.
Until recently, this was mainly a staffing question. Whoever had more service staff, with more experience, could answer more questions more quickly. Waiting time and knowledge largely tracked headcount. That relationship is now under pressure.
Behind every service contact lies a series of actions. The first is triage: what is the problem, how urgent is it, who should pick it up. The second is looking things up: what information about this product, this customer, this installation history is relevant. The third is diagnosis: is this a known pattern or something new. The fourth is the solution itself, and the fifth is recording what happened so that next time goes faster.
Of these five steps, triage is often already possible to automate: a system that categorizes and routes incoming reports does so consistently and without waiting time. Looking things up is a task where AI with human oversight is increasingly taking work off people's hands: searching manuals, previous tickets and maintenance logs can be done faster by a system than by a person, but the translation into what needs to happen in this specific case is still assessed by someone who approves or rejects it. Diagnosis for known patterns is shifting in the same direction. Carrying out the solution, especially if it is physical, and reassuring the customer in a difficult conversation, remains human work.
If looking things up and part of the diagnosis are done by a system, what is distinctive changes. Response time used to be a function of how many people were on the line; if a system categorizes reports immediately and already has the right information ready for the employee handling the conversation, response time becomes less dependent on staffing and more on how well that system is set up. A company with fewer people on the phone can thus respond faster than a company with more people but no supporting system.
Knowledge, too, is shifting from something that lives in people's heads to something that is findable. In the past, an experienced technician who knew every installation by name was an advantage that could not be matched. That advantage still exists for exceptional cases, but for recognizable, recurring problems — the largest part of the reports at most companies — the question is no longer who has the most years of experience, but who has recorded and made searchable the previous cases best. A supplier that has only existed for two years but has structured its service history well can be faster on this point than a supplier with thirty years of experience and unstructured files.
This shift is not happening at the same pace everywhere, and that is not primarily due to the size of the company. It is due to the state of the underlying data. A company whose maintenance logs, ticket history and customer communication sit in separate systems that do not talk to each other cannot feed an AI system what it needs, however advanced that system may be. A company that does have this in order sees triage and looking things up shift faster, while a comparable company in the same sector remains almost entirely dependent on people on this point. That difference is measurable in waiting time, in first-contact resolution time, and in whether a customer has to tell his story again every time he asks a repeat question.
This dimension does not operate independently of the other axes on which companies distinguish themselves. How broadly a supplier can expand its services without hiring proportionally more people is connected to what happens here, as is how AI changes competition on product breadth. The question of how deep the technical knowledge is that a customer gets on the phone also touches on what is at play here, as described in how AI changes competition on technical knowledge. And availability — whether a customer reaches someone or something useful at any moment — is in effect the other side of the same coin, elaborated on in how AI changes competition on availability.
This is not an argument for replacing service staff. What role AI takes on in the service organization and what that means for the workforce is a decision for the employer, subject to its own statutory requirements. What is described here is a shift in what a customer experiences and in what distinguishes suppliers, not a statement about who should keep doing what work.
If your after-sales service is slower or less consistent than a competitor's, the first question is not how many people need to be added, but which part of the work gets stuck on missing or fragmented information. How to address that lag depends on where in the process the problem lies, which is covered in what to do about a lag on a dimension.
If you claim to respond faster or better than your competitors, whether that claim holds up under scrutiny is something to first get clear for yourself, as elaborated on in how to substantiate a claim about your own quality. The free dimension check helps with that: you name where you believe you win on after-sales service, and see which of those claims can be defended with evidence. The full benchmark, with your position and that of your peer group on all dimensions, is under construction.
The underlying question — which part of the work behind your after-sales service can already be taken over by AI today, and which part remains human work — is answered per task with the work scan from FTE TO AI.