From Data Chaos to Transparency: How We Built Analytics for Rostelecom

Key Project Stages

• Stage 1: Preparation Phase
  • Platform identification (Yandex Cloud)
  • Customer requirements analysis
  • Selection of tools (Yandex.Metrica, AppMetrica, ClickHouse, DataLens)

• Stage 2. Setting up data collection

  2.1. Implementing web analytics

  • Installed Yandex.Metrica on 2 websites:
    • Welcome landing page
    • Main platform with courses

  2.2. Setting up mobile analytics

  • Integrated AppMetrica tracking into iOS and Android apps

  2.3. Configuring conversion tracking in Yandex.Metrica

  • Transitions to introductory lessons
  • Transitions to registration
  • Platform registration
  • Lesson views on the platform

• Stage 3. Data integration and processing

  3.1. Setting up user event parameter transmission

  • Each user registered on the platform is assigned a unique user ID (UserID) ➡ We pass these to Yandex.Metrica as visit parameters ➡ Link them with Yandex.Metrica’s ClientID

  3.2. Organising data storage

  • Created a database using Managed Service for ClickHouse within Yandex Cloud

  3.3. Automating data transfer

  • Developed Python scripts for:
    • Advertising costs (Yandex.Direct, VK)
    • Data from Yandex.Metrica
    • Data from AppMetrica

• Stage 4. Infrastructure setup
  • Creating a virtual machine via Compute Cloud
  • Uploading data transfer scripts and configuring their regular execution

• Stage 5. Data visualisation
  • Configured data visualisation in DataLens so the client receives the final analysis results in the form of a dashboard.
  • During the project, we created:
    • Datasets and data merging at the dataset level
    • Connection to ClickHouse
    • Charts
    • Dashboards