if i need to directly compared retention rates of 2 cohorts, but i need a percentage change, how do i compute this? eg: cohort a has a higher daily retention rate tan cohort b by 20% how to compute this 20%?
There are a couple of ways to compute the percentage change between two cohorts' retention rates in Mixpanel:
1. Using the "Compare to Segment" Feature in Retention Reports
In the Retention report, you can break down by cohort and use the "Percent Change over Baseline" comparison option. This lets you set one cohort (e.g., Cohort B) as the baseline and see the percentage lift of the other cohort (Cohort A) against it. [Comparisons]
Specifically, for the Retention Curve chart type, "Percent Change over Baseline" is supported, which would give you exactly the kind of relative difference you're describing. [Comparisons]
2. Manual Calculation
The percentage change is computed as:
% Change = ((Cohort A Retention Rate - Cohort B Retention Rate) / Cohort B Retention Rate) × 100For example, if Cohort A has a Day 1 retention of 50% and Cohort B has 41.7%:
% Change = ((50 - 41.7) / 41.7) × 100 ≈ 20%This is the same formula Mixpanel uses for lift calculations in its Experiments feature. [Experiments lift]
Practical Steps in Mixpanel
Go to the Retention report.
Add your two cohorts as a breakdown.
Use the Compare → Percent Change over Baseline option, setting Cohort B as the baseline.
This will surface the relative percentage difference in retention rates between the two cohorts directly in the UI, without needing to calculate manually. [Comparisons]
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
kapa.ai what if it is more like "users in cohort a" vs "users NOT in cohort a"
The same approach works for this case. When you break down a Retention report by a cohort, Mixpanel automatically shows both groups — users in the cohort and users not in the cohort. [Breakdowns]
So the steps are identical:
Go to the Retention report.
Click Breakdown and select your cohort (e.g., "Cohort A").
Mixpanel will split results into two segments: "In Cohort A" and "Not In Cohort A".
Use Compare → Percent Change over Baseline, setting "Not In Cohort A" as your baseline.
This will directly show you the percentage lift of users in Cohort A relative to those not in it — exactly the "20% higher retention" type of comparison you described.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
kapa.ai can i get an average percentage across all days/weeks/months, based on if im looking at weekly, monthly or daily retention?
Based on the knowledge sources, Mixpanel does compute an average retention rate across all cohort buckets — this is the "Average" row shown at the top of the Retention report table.
Here's how it works:
Mixpanel calculates the Average row by taking a weighted average of all completed buckets, weighted by the number of users who entered each cohort. Incomplete buckets (still in progress) are excluded from this calculation. [Retention Average Row]
This applies whether you're looking at daily, weekly, or monthly retention — the same weighted average logic is used across all time units. [Retention FAQ]
So if you're comparing "In Cohort A" vs "Not In Cohort A" and want a single average percentage across all days/weeks/months:
Look at the Average row for each segment in the retention table.
Apply the same percentage change formula discussed earlier:% Change = ((Cohort A Avg Rate - Not Cohort A Avg Rate) / Not Cohort A Avg Rate) × 100
However, I should note that the knowledge sources don't explicitly confirm whether the "Percent Change over Baseline" comparison option in the UI applies to the Average row specifically. You may need to read the Average row values manually and compute the percentage difference yourself if the UI doesn't surface it directly.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
how do i get the first album users listen to when they open the app? i have an event named "Album_Opened" so how does this work? kapa.ai
You can use the "First Time Filter" on your Album_Opened event to identify the first album a user listens to when they open the app. Here's how it works:
Using the First Time Filter
In any Mixpanel report (Insights, Funnels, Retention, or Flows), you can add a First Time Ever filter to your Album_Opened event. This filters the event down to only the first instance of that event performed by each user. [First Time Filter]
Mixpanel computes this on-the-fly by scanning each user's history to determine if a given Album_Opened event was the first time they performed it, based on timestamp. [First Time Filter]
Practical Steps
Go to Insights (or Funnels/Flows depending on your goal).
Add the Album_Opened event.
Click the … icon next to the event and add a First Time Filter.
Breakdown by the album name property (e.g., Album_Name or whatever property you use to identify the album).
This will show you the distribution of which albums users opened for the very first time, effectively answering "what is the first album users listen to when they open the app?"
Additional Options
You can also create a cohort of users who opened a specific album for the first time, using the First Time Filter in the Cohort Builder. [Cohorts]
If you want to see the path leading up to or following the first Album_Opened event, you can use the Flows report with the First Time Filter applied. [Flows]
Note: If two users open an album at the exact same timestamp, all those events will be considered "first time" events. This edge case is unlikely but worth being aware of. [First Time Filter]
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
how do i get the first album users listen to whenever they open the app, like per day/per session, what album do they start with? i have an event named "Album_Opened" so how does this work? kapa.ai
This is a slightly different question from the previous one — you're not looking for the first-ever album a user opens, but rather the first album they open per day or per session (i.e., the "entry point" album for each day/session).
Here's what the knowledge sources support:
Per Session: Use Sessions in Funnels or Flows
You can use Sessions in Mixpanel to scope analysis within a single session. In the Funnels report, once sessions are configured, a "Session Start" and "Session End" event are generated. You could build a funnel like:
Session Start → Album_Opened
Then filter or breakdown by the album name property on the Album_Opened step. This would show you which album users open first within a session. [Session Controls]
In Flows, you can also use "Session Start" as an anchor event and see what events (including Album_Opened) follow it, weighted by number of sessions. [Session Controls]
Per Day: Breakdown by Date in Insights
For a per-day view, you can use Insights with the Album_Opened event and:
Add a First Time Filter to Album_Opened — but note this gives you the first ever, not first per day.
Alternatively, breakdown by Day of Week or Date to see which albums are opened most frequently on a given day.
However, the knowledge sources do not explicitly describe a native "first event per day per user" computation in the Mixpanel UI. The closest supported option for per-day first-event analysis would be via JQL (JavaScript Query Language), where groupByUser() can compute, for each user and each day, their first event on that day. [JQL groupByUser]
Summary
Goal | Recommended Approach |
First album per session | Funnels or Flows with Session Start → Album_Opened |
First album per day (advanced) | JQL with groupByUser() keyed by user + day |
Most common "entry" album overall | Insights + breakdown by album name property |
I'd recommend starting with the Funnels + Sessions approach for per-session analysis, as it's the most straightforward UI-based method. [Session Controls]
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.