You know the service well, but a discovery call can still become a presentation before the client's problem is clear. Practising the conversation can reveal that habit before a real enquiry is involved.
Use AI role-play as a rehearsal with a fictional client situation, defined learning objective and structured debrief. The simulated answers are practice material. They are not customer research or evidence that your intended market thinks a particular way.
Photo by LinkedIn Sales Solutions on Unsplash.
Choose one skill to practise
A useful exercise might focus on asking about the current process, clarifying consequences or checking whether the person has authority to make a change. Avoid asking the assistant to play every kind of client in one conversation.
Write the objective in observable terms. For example: identify the current handoff and its main interruption before explaining a possible service. That is easier to review than become better at sales.
Use a completely fictional scenario rather than copying a client's confidential conversation into an unapproved tool.
Give the role enough context
Prepare a small scenario card:
Fictional organisation and role.
A recognisable operating problem.
Information the character can reveal when asked.
One uncertainty or competing concern.
A boundary on what the role should not invent.
The behaviour the learner is practising.
Avoid building stereotypes into the character. A difficult conversation can come from unclear information or conflicting priorities without attributing behaviour to personal characteristics.
An illustrative founder scenario
The fictional founder says the team needs automation. The hidden detail is that two staff members use different approval rules, so the founder keeps checking every result.
The learner's task is to ask how the work currently flows and what happens when cases differ. A premature tool recommendation misses the useful discovery. The role-play should reveal the detail only when the question makes it relevant.
This scenario is designed for practice. It does not establish how frequently that problem appears in Hili's actual enquiries.
Debrief the questions, not the performance theatre
After the exercise, review which question produced useful information, which assumption went untested and where the learner started pitching too soon.
Ask for examples from the conversation rather than a vague confidence score. The learner should be able to see the wording that helped or hindered the discussion.
A tool's feedback is another suggestion to consider. A knowledgeable colleague can review the exercise where commercial judgement or service positioning matters.
Make the next rehearsal different
Change one important feature, such as the source of the delay or the person's ability to decide. This checks whether the learner can use the approach without memorising a script.
End with a few questions worth taking into a real conversation. Keep them flexible so listening remains possible. The purpose is to prepare for uncertainty, not train people to push every client through the same predetermined diagnosis.
For related practical guidance, see deciding when operations support is useful and asking practical implementation questions.
Hili Consulting can help design a fictional role-play that improves discovery questions. Arrange a free 30-minute consultation and bring one example of the work you want to improve.
