#7 Intro to the RetentionKit package

I am happy to share the first product demo from DataDiner.

This is RetentionKit. In my newest YouTube video I set out to answer the following questions: Do you have product-market fit? How healthy is your product growth right now? And when metrics drop, how does that affect retention — which cohort took the hit, and what caused it?

The mission of DataDiner has always been to make data analysis accessible to anyone — so RetentionKit is open source, and ships with an AI data analyst (Claude Code based for now) that runs the analysis for you. Clone the repo and point it at your own data, or just follow the guided exercise in the video and answer the questions with me.

What is in the video?

So in this video, I introduce RetentionKit, a Python-based toolkit for analyzing product retention — combined with an agentic AI system that lets you explore your data through natural language.

You’ll see how to:

• Set up and explore the RetentionKit repository
• Prepare event-based product data for retention analysis
• Define your product’s core value event
• Handle sparse event data and create continuous user timelines
• Generate key retention visualizations, including:
  • Usage frequency histograms
  • Retention curves
  • Lifecycle and Quick Ratio charts
  • Cohort retention heatmaps
• Use AI to explore the repository and understand the analysis workflow
• Run a complete retention analysis on a sample chess platform dataset
• Investigate retention trends and user segments

The goal of RetentionKit is to make advanced retention analysis more accessible — including for professionals who don't have a Python background.

The project is still in its first version, with additional charts, workflows, and integrations planned for the future.

It's v1 — tell me what you think!