I have 2 cohorts, but the audience overlaps - how do I make sure there's only one user in the funnel when I breakdown by cohorts? I want to use the first touch attribution.
This is a nuanced situation. Let me address both parts of your question:
Overlapping Cohorts in Funnel Breakdowns
When you break down a funnel by cohorts, Mixpanel counts users in each cohort segment they qualify for. This means a user who belongs to both cohorts will appear in both segments โ there is no built-in mechanism to force a user to be counted in only one cohort when breaking down by cohorts. [Funnels Breakdowns]
The knowledge sources do not explicitly describe a way to deduplicate users across overlapping cohort segments in a funnel breakdown.
First Touch Attribution in Funnels
For attribution, you can use the "Attributed byโฆ" breakdown in Funnels, which attributes credit for the funnel to touchpoints along the user journey using the first step of the funnel. You can select First Touch as your attribution model, which gives 100% credit to the first touchpoint within the attribution lookback window. [Attributed by Breakdown; Attribution Models]
To set this up:
Add your conversion metric to the funnel.
In the Breakdown section, choose the "Attributed byโฆ" computed property.
Select the property you want to break down by (e.g., UTM medium).
Change the attribution model to First Touch in the metric section.
โ ๏ธ Note: Attribution is available on Enterprise or Growth plans only. [Attribution]
Summary
Overlapping cohorts: There is no native way in Mixpanel to ensure a user is counted in only one cohort segment when both apply โ the knowledge sources don't describe such a deduplication option for cohort breakdowns.
First touch attribution: This is supported via the "Attributed byโฆ" breakdown with the First Touch model, but this applies to property-based attribution (e.g., UTM channels), not cohort-based segmentation.
If your goal is to attribute users to a single channel/touchpoint, the "Attributed byโฆ" breakdown with First Touch is the right tool. If you need mutually exclusive cohort segments, you would need to redefine your cohorts so they don't overlap before applying the breakdown.
โน๏ธย Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
