CASO DE ÉXITO
Incrementing Customer Lifetime Value
Selecting the group of clients to which address a churn prevention campaign is a key aspect to success, to boost client fidelity and also to reduce the costs of campaigns. Our Machine Learning approach at CleverData has demonstrated that:
- The combination of traditional BI and new Artificial Intelligence approaches helps in solving both classical and new problems.
- Sorli is now able not only to predict which customers will churn, but also to create and follow their campaigns with easy to use and intuitive Dashboards.
«Marketing campaigns now use client segmentations based on behavioral data, not on static rules. Thanks to periodic retraining, predictive models follow and adapt to changes in client behavior»
As part of their path to become a data-driven company, Sorli, with hundred of thousand clients loyalty card customers, wanted to use advanced techniques to segment their clients based in their behavior more than on static rules, as they were traditionally doing.
They also wanted to upgrade the manual management of their marketing campaigns into practical and easy to use dashboard
HOW PRODUCT HELPED
Predictive analytics let Sorli target the clients they are really interested in. Marketing campaigns now use client segmentations based on data, not on static rules. Thanks to periodic retraining, predictive models follow and adapt to changes in client behavior.
No more spreadsheets to create and follow campaigns. Our brand-new visual dashboards let Sorli control campaigns success, ROI, and several KPIs, comparing results with control groups or past campaigns. Machine Learning let them extract value and make automated decisions based on data.
RESULTS AND FUTURE PLANS
Results have shown that customers detected by our predictive model after the promotions have raised 15% of increment in their average basket and a 28% in their frequency resulting in a 47% of the total amount.
Future plans rely in put in play whether forecast and competence promotions in order to improve models.
«Fast prototyping makes it easy to test and build MVPs (Minimal Viable Products) that create the needed trust to build more sophisticated products.»