Case Study · General Insurance · Pan-India
100%
Of support calls audited automatically, zero manual sampling
80%
Lower cost to audit every interaction
12K+
Support calls handled daily, pan-India operations
Client Profile
A private general insurance company in India, still a startup by size but running sales and support nationwide. 60 agents on sales calls, another 60 on support, no single city or language covering the whole book of business. Policy issuance and claims were already digitized, so the paperwork problem was solved. What remained was the phone.
12,000 support calls a day, and average handling time sitting between 5 and 7 minutes. Every one of those calls was supposed to be reviewed for quality, but review meant a person listening back to a sample, by hand, after the fact, in whichever language they happened to speak. There was no way to tell if an agent in one city was being held to the same standard as one in another, and nothing built to catch a call that came in in a regional language rather than English.
After CanString, average handling time came down to 5 minutes. 100% of calls got audited automatically, not a sample. Audit cost dropped 80%. The QA layer that used to catch a fraction of what agents said now catches all of it, in whatever language it was said in.
The Situation
Before
Support and sales handled 12,000 calls a day across regional languages, with average handling time between 5 and 7 minutes. Audits were manual, slow and covered only a small sample, with no way to track consistency across agents.
After
Analyzer and BI run as one layer across voice and non-voice channels, in the agent's own language. Every call is scored against the same rubric automatically, driving ACHT down and QA score up.
Free Download
Get the full case study
The full QA picture, the multilingual analytics approach, and the rollout timeline, sent straight to your inbox.
- 7-page PDF, no fluff
- Read in under 5 minutes