The Turkish auto market has been highly competitive in recent years, with the added pressures of volatile economic inflation and regular changes in regulations. Increasing pricing sophistication and a wave of new market entrants focused on gaining market share have further driven the need for, and benefits of, strong technical and retail pricing—particularly in the Casco (vehicle damage) segment of the market.
Faced with this combination of competitive pressures, AXA Sigorta, one of the market leaders, determined a need to strengthen technical pricing to improve margins, volumes, and the stability of its auto business.
Among the targets for the pricing transformation work that began in 2015 has been an improvement in Casco customer renewal/retention rates, a project for which the company decided to request competitive bids.
WTW has supported AXA Sigorta throughout its pricing transformation journey, beginning with the development of a technical pricing roadmap to position it as a market leader based on the use of our Radar software suite, including Emblem, Classifier, and Radar Base.
After progressing through projects on that roadmap, the company’s attention turned to improving Casco auto product retention rates without impacting its combined operating ratio (CoR) and accelerating its pricing sophistication using optimization and machine learning in a dynamic pricing environment.
AXA Sigorta sought a solution that provided robust modeling capabilities to meet its current needs and also a platform to create and use machine learning and artificial intelligence, which are key parts of its pricing strategy.
Having been successful in winning the project in competition with other software/service providers, our team started by understanding AXA Sigorta’s existing retention processes to build retention and cancellation models following the customer journey. The addition of the Radar Optimizer module to AXA Sigorta’s Radar package enabled this to be integrated into a price optimization environment that uses customer demand models to indicate propensity to renew at different price points. A/B testing was used in the implementation phase ahead of a full rollout of renewal price optimization in 2021.
“The combination of WTW’s software and technical pricing expertise has helped us develop a customer retention optimization model that we believe really strengthens our position in the highly competitive Turkish Casco market.”
Sanem Çınay Buçukoğlu
Chief Operating Officer, AXA Sigorta
“The combination of WTW’s software and technical pricing expertise has helped us develop a customer retention optimization model that we believe really strengthens our position in the highly competitive Turkish Casco market.”
Sanem Çınay Buçukoğlu | Chief Operating Officer, AXA Sigorta
The project went through a monitoring phase comparing a 30% non-optimized portion of the portfolio with a 70% optimized portfolio, in which the margins for the different segments were kept consistent and a focus was on reducing lapses. The lapses in the optimized cohort were 5–10% lower than the lapses in the non-optimized cohort.
With these signs of success, AXA Sigorta has been confident to continue with further phases of its pricing transformation roadmap, which includes embedding the use of the gradient boosting machine (GBM) machine learning techniques available in Radar in its pricing process.
