How to Beat Chaos in User Behavior Data

The analysis of user behaviour in large digital products often comes up against the problems of working with raw data with a complex structure, the lack of a unified attribution scheme for actions, and the fragmentation of expertise. This leads to low efficiency of analysts, the risk of making wrong decisions based on disparate metrics, and missed cross-functional opportunities.
In this talk, we will present an approach to solving these problems by building a centralised user action mart. We will explain how we standardised the collection, attribution, and representation of user actions (views, clicks, add-to-cart events) linking them to the business context. You will learn how this approach has simplified the data, unified the methodology for calculating key metrics for A/B tests and dashboards, and freed up many hours of analysts' work.
The talk will be useful for system and business analysts and for all analysts, heads of analytical departments, data architects and product managers.