We need an AI chatbot. Or do we?
The idea seemed clear: An AI chatbot should reduce the workload in customer service. But a closer look at the actual workflow revealed that the real problem was somewhere else.
Sometimes, the starting point is exactly the opposite: There is no shortage of ideas. In fact, the solution already seems obvious.
That was the situation for the Head of Customer Service at a medium-sized mechanical engineering company. His team handles numerous technical customer inquiries every day. Many of them are similar, yet processing them takes a considerable amount of time.
So when he started with the Use Case Pilot, his idea was already quite specific:
“We need an AI chatbot for our customer service.”
He actually wanted to find out how this idea could be implemented.
But before discussing the chatbot, the Use Case Pilot asked a different question: What problem is it actually supposed to solve?
The solution seemed obvious
A concrete idea is a good starting point. But it is still worth understanding the problem behind it first.
Challenge / starting point
Reduce the workload in customer service and speed up the handling of technical customer inquiries.
Experience with the Use Case Pilot
At first glance, the connection seemed obvious:
Lots of customer inquiries plus a busy service team equals a chatbot.
But the Use Case Pilot took a step back and explored how a typical customer inquiry is actually handled today.
What happens after an inquiry comes in? Where does the effort occur? What information does a service employee need before being able to respond?
It became clear that writing the actual response was often not the biggest problem.
Result
Instead of immediately developing a solution that already seems obvious, the user first identifies where the greatest potential for improvement lies in the actual workflow.
The real problem came before the answer
Communicating with the customer was not taking most of the time. Finding the right information was.
Challenge / starting point
Understand which steps in handling a technical customer inquiry actually require the most time and effort.
Experience with the Use Case Pilot
A closer look revealed what service employees have to do when a new inquiry arrives.
First, they need to identify the machine and its specific configuration. Then they search technical manuals, documentation, and previous service cases. Only after gathering and assessing this information can they prepare a well-founded response.
This changed the question.
Instead of:
“How can we answer customer inquiries automatically?”
The first question became:
“How can our service employees get the right information for this specific case faster?”
Result
The focus shifts from an obvious technical solution to the part of the workflow that actually creates a significant share of the workload.
A different first use case
The chatbot was not necessarily a bad idea. But it might not have been the best first step.
Challenge / starting point
Give service employees faster access to a solid information base for handling technical customer inquiries.
Experience with the Use Case Pilot
Looking more closely at the workflow led to a different use case:
An internal service assistant could bring together information about the specific machine, its configuration, and its service history with relevant technical documentation.
When a new inquiry arrives, the service employee could receive the information that is most relevant to that particular case.
The employee reviews the information, assesses the case, and decides what response the customer should receive.
The original idea of a customer-facing chatbot does not have to be discarded. It could still be useful.
But it may be step two rather than step one.
Result
The service team receives support exactly where a significant part of the work occurs today, while the professional assessment and final decision remain with the service employee.
A solution idea is a starting point, not the result
Even a fairly concrete idea does not necessarily mean that you have found the right use case.
“We need a chatbot” initially describes a possible solution. Only by looking at the actual workflow can you understand which problem needs to be solved and where support could make the biggest difference.
In this case, that closer look led to a different first use case.
The original idea was not necessarily wrong. It may simply have come too early.
Bring your solution idea to the Use Case Pilot. You do not need to know yet whether it is actually the right one.
Discover your own AI use cases
Start with the Use Case Finder and develop concrete use cases for your own work context.
Open Use Case Finder