Close this search box.

Increasing insurance customers’ lifetime value with predictive upsell analytics


from diagnostics to model deployment and first results


of upsell success rate after the first 6 months


of customers likely to buy more
are precisely identified due to the high accuracy of the model​

The business problem

A major British insurance company was facing a less than satisfactory upsell rate for their automotive insurance customers.

Our task

We were tasked to come up with a model to identify which customers are likely to accept and upsell offer and suggest what offers would be most relevant to customers.

Our Solution

We created a model that predicts how likely is each customer to accept an offer for complementary products and services – each of them is assigned a “propensity to buy” score from 0% to 100%. The model is based on customer, transactional, service and NPS® survey data. After initial validation the model was implemented into the customer service processes and started generated regular recommendations, optimisation suggestions, more accurate prioritisation of customer segments.

A look in the future

After the first 6 months of targeted marketing initiatives for customers with a high likelihood to buy, upsell success rate increased with 28%.

Project Trivia:

Industry: Insurance

Company Size: 4500+ employees

Location: Switzerland

GemSeek Capabilities: Predictive Customer Analytics

Retain, Advance and Grow your customers with predictive customer analytics.

Related posts