Turn raw Cursor usage data into defensible adoption cohorts, a reusable analysis, and practical coaching that helps each group try the behavior most likely to deepen its use of AI.
Source episode · 20:00
How I AI: Chintan Turakhia’s Playbook for AI Adoption at Coinbasewith Claire Vo
Chintan exports Cursor admin analytics, asks Cursor to find natural usage cohorts, generates a reusable Python analysis and HTML dashboard, then turns the patterns into guidance for moving each cohort forward.
A reproducible cohort analysis, visual dashboard, and role-appropriate adoption playbook grounded in actual usage patterns.
AI-first code editor
Anthropic AI assistant
Step by step
Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.
5 steps
Export the team analytics CSV from Cursor and inspect its schema. Remove fields you do not need, protect identities, and define the behaviors you want to understand before asking for clusters.
The CSV file typically contains fields like `accepted_lines`, `chat_lines`, and other interaction metrics that are crucial for the analysis.
Ask Cursor to profile the data and find natural usage groups using volume, agent versus tab behavior, acceptance, feature breadth, and model choices. Require assumptions, group sizes, and a reusable script.
Analyze this Cursor admin export to find natural usage cohorts. Use accepted lines, chat activity, agent or composer activity, tab activity, acceptance behavior, feature breadth, and model usage where available. Explain preprocessing and assumptions, test several reasonable cohort counts, label groups by observed behavior rather than value judgments, and write a reusable Python script plus enriched CSV.
Read the plan before execution. Confirm the proposed metrics, privacy handling, clustering method, outputs, and limitations match the enablement question rather than employee performance measurement.
The AI's plan will likely include steps to identify cohorts, generate a Python script for analysis, and create an HTML dashboard for visualization.
Run the analysis and generate a static HTML dashboard with cohort sizes, defining behaviors, distributions, and representative aggregate patterns. Keep the underlying script and output reproducible for the next export.
The dashboard can help you quickly identify different user personas, such as Agent-Heavy users vs. Tab-Heavy users.
Ask for one practical next behavior per cohort and a clear path between behavior patterns. Turn the result into a short playbook and Slack summary, then check every recommendation against the data.
Using the validated cohort analysis, create guidance for each cohort. Name the observed behavior pattern, one useful next experiment, an example task or prompt, and the metric that would show progress. Do not rank employee performance or claim productivity from generated line counts. Produce a concise HTML playbook and a short Slack summary.
Turn an idea into a PRD, user stories, and a plan.
After the steps
How to recover when the loop fails and where human judgment helps.
Recover
Keep building
Log your workouts effortlessly without ever opening a spreadsheet. This workflow uses the Claude mobile app to parse a simple voice note and automatically update your workout tracking spreadsheet in the background.
Create a custom email management interface within Claude to read, triage, and draft replies without opening the Gmail tab. This workflow helps you reclaim focus and compound your learnings by training your AI on every interaction.
Build a fully automated system in Claude to generate personalized, interactive HTML proposals and client onboarding materials, saving hours on administrative work and creating a high-touch experience.
Join 100,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.