From 60 to 10 minutes: one assessment pilot in numbers
How a European IoT security company assessed its support team, and what the managers saw for the first time.
The company runs 24/7 support for partners, installers and end users in more than 180 countries. A manual evaluation of one support agent took about 60 minutes and depended on the judgment of the evaluator. There was no consistent way to compare agents.
The pilot
Eleven support agents talked to AI customers in realistic situations: angry customers, technical complaints, de-escalation under pressure. Each agent was assessed against the same competency framework, with no scheduling and no disruption to the shifts.
- Assessment time: from 60 to 10 minutes for each person.
- 89% of the agents had critical skill gaps that manual review did not show.
- Four competencies measured: problem solving, tension reduction, clarity and politeness, support and trust.
The whole team was assessed in the time that two manual evaluations took before.
What happened next
The company got its first objective and comparable skill baseline for the support team. L&D focused on the gaps that the data showed: paraphrasing, response structure and trust-building. The name of the company is anonymised.
SoftTrainer team