Analyzing Correlation Between User Properties and Other Values
how can I do an anlysis on correlation between a set of properties in users with another value? what would you call that and which combination of reports we can use?
how can I do an anlysis on correlation between a set of properties in users with another value? what would you call that and which combination of reports we can use?
Hi Juan I., Please stand by as I search the Mixpanel knowledge base for an AI-powered answer. I normally take about 15 seconds or so. In case AI cannot answer your question, feel free to reach out to Mixpanel support at mixpanel.com/get-support. If you have follow-up questions for this thread, please tag me again with your question: @kapa.ai <follow-up question> 👋
Feel free to ask a follow-up question once I've answered by mentioning me with @kapa.ai
To analyze the correlation between a set of user properties (or behaviors) and another value (such as a conversion event, retention, or another property), Mixpanel provides several tools and report types. Here’s how you can approach this:
1. Correlation Analysis (Signal Report)
The most direct way to analyze the correlation between user behaviors (events) and a goal (like retention or conversion) is with the Signal report. Signal quantifies the association between a "correlation event" (what users do) and a "goal event" (the outcome you care about). It uses statistical measures (phi coefficient) to show how strongly certain actions are associated with your goal, and provides actionable insights such as the optimal frequency and timing of those actions.
- Use Case: Find out which user actions (or event properties) are most strongly correlated with a desired outcome (e.g., retention, purchase). - How to Use: Select your target users, correlation events (with optional property breakdowns or filters), and your goal event. Signal will return the strength of correlation and key findings. - Limitation: Signal currently supports filtering and breakdowns by event properties, not user properties directly. If you want to analyze by user properties, you may need to create cohorts first or use breakdowns in other reports. - What it’s called: Correlation analysis or behavioral correlation analysis.
Docs: Signal Report Reference: Calculations & Interpreting Results
2. Cohorts and Breakdowns
If you want to analyze how different user property segments (e.g., country, plan type) relate to another value (like conversion rate), you can:
- Create Cohorts: Group users by specific properties or behaviors (e.g., "Users from US", "Power Users"). - Breakdown Reports: Use the Breakdown feature in Insights, Funnels, or Retention reports to group results by a user property or cohort. This lets you see how different segments perform on your key metrics. - What it’s called: Segmentation analysis or cohort analysis.
3. Find Interesting Segments
In Funnels and Retention reports, the Find Interesting Segments feature automatically identifies property segments or cohorts that significantly outperform or underperform on your chosen metric (e.g., conversion or retention rate). This helps you discover which user segments are driving or hindering your outcomes.
Docs: Find Interesting Segments
Summary Table:
Note: If you want to analyze correlation between user properties (not event properties) and another value, you may need to first create cohorts based on those user properties, then use breakdowns or the Find Interesting Segments feature in your analysis. Direct correlation analysis between user properties and outcomes is not explicitly supported in Signal as of now.
If you need to analyze correlations between time-series metrics, the Metric Tree also offers a Pearson correlation coefficient between metrics, but this is more for metric-to-metric relationships rather than user-level property analysis Metric Tree Correlation.
If you have a more specific scenario or data structure, let me know and I can suggest the best combination of reports for your use case!
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.