Data from Yandex.Metrica can be used to not only analyze user behavior after the fact, but also predict subsequent user actions. Yandex.Metrica, along with online DIY retailer “220 Volt”, has built a model that calculates the conversion probability for each site user. Head of Yandex.Metrica Viktor Tarnavsky will talk about the steps needed to create a predictive analytics model at the eMetrics summit in London on October 11-12, and in Berlin on November 13-14.
This predictive model takes in raw data from Yandex.Metrica and processes it using machine learning algorithms. The model then assigns every user a rating according to their likelihood of converting. You can build remarketing lists using these ratings and address different groups of potential clients more effectively.
“Machine learning” might sound like rocket science – but don't be daunted! There are ready-made machine learning libraries out there, so all you have to do is connect them to your analytics environment. Viktor will share a step-by-step guide that will help you build a predictive model yourself.
During both events, Yandex.Metrica team members will be present at our stands to answer any questions you might have, or simply have a friendly chat on all things analytics.
To make both events even more exciting, please use the promo code YANDEX15 that gives you a 15% discount on all conference participation options excluding networking.
See you in Europe!