Use case

Personalized offers recommendation

Problem definition

Customers are nowadays overwhelmed with too many notifications about different products and services, and often confused by all the information. Helping them select the best offers from an active portfolio improves their satisfaction and increases product revenue. The portfolio development usually depends on company focus, verticals and product maturity.


With a fully automatized contextual marketing recommendation engine, the company can improve all of its manually managed CRM campaigns, increasing offers acceptance rate up to 30% and product revenue up to more than 3%. In line with the omnichannel approach, the most relevant offers are available for every customer on all communication channels. This helps retail agents present the best offers for customers at any touchpoint.


The automatized engine combines several machine learning algorithms to produce relevant offers at the right moment, for the right customer. The engine is fully integrated with the campaign management system, as well as front-end channels for real-time recommendations. The engine consists of propensity to buy models, clustering models and price level estimations, considering channel preferences. The results are analyzed against profitability and contact policy rules.


The relevant customer segment is targeted with the best offers from the product portfolio, while the results are being measured on a daily level and presented on a customized dashboard.