A large international insurance firm had been conducting a customer survey for years. But because the survey had been designed to elicit responses as unstructured data, our client had no simple way to analyse it. Instead, the insurer manually reviewed a small number of comments to understand the ‘voice’ of its customers.
In short, the insurer derived little or no value from the data and the survey – until we introduced text analytics to the business.
How we helped:
We designed a process to manage the data flows from the survey provider through an analytics engine to produce a wealth of reports for the frontline teams. We also classified the responses into more than 80 categories, reflecting different questions the business wanted to answer. The reports we generated gave the insurer a rich understanding of what drove their customer sentiment – as well as a host of product or service-specific insights.
We then worked alongside the client’s training teams to implement a structured coaching programme providing targeted actions that reflected the insight.
The content of the unstructured data was classified and understood across 100% of the insurer’s customer feedback, which enabled it to use the data in different ways. As well as underpinning the coaching, the insurer’s marketing and product teams were able to understand which aspects of their products had the greatest impact on customer sentiment, helping them to prioritise improvements.
Together, these actions have resulted in a 10-point improvement in Net Promoter Score (NPS) for the business, through agent coaching and product and process improvements.