PREDICTING INDIVIDUAL CUSTOMER RENEWALS Next Case Study

Customer:

A Leading General Insurance Company In India

Business Background:

Motor vehicle insurance policies, especially in segments like tractors, have abysmally low renewal rates in India. Individual customer decisions on renewal may be affected by a variety of factors such as location, relationship with dealer, quality of service from the company in past year, claims filed and paid, pricing for renewal, market share in the given geography, etc. Identifying in advance customers who are most likely to renew helps in focusing field efforts for maximum ROI.

Need For Analytics:

Hundreds of thousands of policies per year, each with scores of attributes make manual examination very hard. Further, renewal behavior varies substantially with the type of the vehicle and customer location, requiring an ensemble of prediction models.

Technical Solution:

Data was extracted from internal CRM and other systems, and blended. Substantial cleanup was required in terms of filling / deleting missing values. New attributes were generated using domain knowledge inputs. In total, there were hundreds of thousands of rows and hundreds of columns.

From relative contributions of individual attributes towards renewal, promising attributes were shortlisted, and multiple model building methods were run on that data. Models were also built at the individual product level and individual geography level. Explicability of patterns was a primary consideration for the customer, so rule & tree based classifiers were preferred. A side benefit of the models was a structured way of adjusting the data capture strategy.

Software / Tools:

R, Knime, Open Refine and WEKA

Outcome:

For multiple geographies and vehicle models, more than 95% accuracy was obtained in predicting renewals. Serious paucity of useful data for certain segments was identified.