The starting point: high volume, low visibility
The insurer had an efficiency problem manifesting as long handling times, stagnant policyholder satisfaction, and high agent turnover. Management knew there was a problem but lacked sufficient data to identify exactly where or how to solve it.
The main bottleneck identified in the diagnosis was post-call time: each call generated 15-25 minutes of documentation work — updating the file, notes to supervisor, internal coordination emails. For 1,200 monthly calls, that was 300-500 hours of monthly administrative work.
Implementation in 3 phases
Phase 1 (months 1-2): Deployment of CallsIQ for automatic transcription of all calls. Integration with the claims CRM for automatic data extraction (policy number, claim type, policyholder data).
Phase 2 (months 2-3): Analysis of the first 500 transcripts to identify patterns in high vs. low satisfaction calls. Development of a new service protocol based on findings. Team training using real transcript fragments as case examples.
Phase 3 (months 4-6): Systematic quality monitoring based on transcripts. Weekly individualized feedback to each handler with examples from their own conversations. Monthly protocol adjustment based on new findings.
Results after 6 months
Average post-call documentation time fell from 20 minutes to 9 minutes — a 55% reduction. Overall operational efficiency improved 40% (more calls handled with the same team). NPS rose from 31 to 49 points, exceeding the 10-point target.
An unexpected benefit: agent turnover dropped from 35% to 18% annually. Agents valued the feedback based on their own real conversations vs. the previous generic coaching.
Critical success factor: Training based on real transcripts (not simulated role-plays) was the element with the most impact on quality improvement. Agents recognize their own communication patterns much more easily when they see them in writing.