Retentioneering

Retentioneering focuses on clickstream and user-path analysis, using event log data to help growth teams, data analysts, product analysts, and developers deeply understand user behavior.

Key Features

• Path Analysis: Paths are the unit of analysis; a raw event log has one row per event. Supports defining complex event sequence patterns like add_to_cart->purchase or add_to_cart->[^support_chat]*->purchase.
• Segment Comparison: Define segments (e.g., device type) via add_segment. Almost every non-trivial analysis is a comparison, and segments define what is being compared.
• Conversion and Churn Analysis: Anchor the analysis on any event, look at what happens right before drop-off, and trace an abstract "conversion fell" down to the concrete screen where the scenario falls apart.
• Transition Graphs and Bottleneck Analysis: Build a transition graph and analyze the main bottlenecks.
• Cluster Analysis: Use metrics as features for clustering and as overview metrics for resulting clusters; clusters can be persisted as segments for diff mode and other segment-aware tools.
• Agent Skills: Teach coding agents like Claude Code or Codex how to write and run retentioneering code directly, without cloning the repository.
• MCP Integration: Provides an MCP server that needs no installation, no kernel, and no data; connect it to any MCP client and the agent can search and read documentation.

Usage

Users can call Retentioneering functions directly via Python code or use Agent Skills to let coding agents generate analysis code automatically. The tool supports running from a repository clone and via MCP integration.