What did we look at?
Customer feedback data set with all open explanations on the NPS, CES, employee and tips/suggestions question from 6 questionnaires from different sectors: pensions, insurance and banking.
How did we do this?
We applied industry-specific analysis to a historical dataset. Based on the tex minin results we looked at the quality of the data and substitutability of closed aspect questions through the topics of text mining.
What were the results?
- The analysis showed that the data quality from the CES, employee and tips/suggestions question is not optimal for text mining. For example, there was a lot of response in the open explanations such as “no opinion” or “no explanation”. Underlined made a new proposal for asking the open question for these 3 KPIs, in order to optimise the response quality.
- The analysis made it clear that all the closed aspect questions posed by the insurer could be replaced by an open question with the topics from text mining. The advantage is that the feedback will be spontaneous instead of being helped. Also, the questionnaire is shortened by taking away the closed aspect questions (such as friendliness, expertise etc.).
What will the company do with the insights?
The insights will be used for continuous steering at department level, team level and employee level. For example, employees gain insight into their key qualities and improvement points from customer feedback.
What are the next steps?
- Update current customer satisfaction surveys.
- Continuously enrich the open explanations with text mining analysis using the Underlined API service and process it in reports.
More customer experiences