Mixpanel + RevenueCat vs one analytics profile: when the glue breaks
Mixpanel and RevenueCat are a legitimate stack. This post covers identity glue, sessionless renewals, and when unified Analytics, Convert, and Engage wins without telling you to delete Mixpanel.
Mixpanel plus RevenueCat is a sensible default for many subscription apps. Mixpanel gives flexible event analytics, cohorts, and funnels. RevenueCat gives billing truth, entitlements, and subscription charts. Thousands of teams run this pairing successfully.
This post is not a hit piece. It explains where the pairing strains: identity glue, renewals without sessions, and the gap between insight and in-app action, and when a unified profile across Analytics, Convert, and Engage reduces operational tax. If Mixpanel is your company-wide analytics standard, you may keep it and still add Rheo for the subscription funnel layer.
What Mixpanel + RevenueCat does well
| Capability | Mixpanel | RevenueCat |
|---|---|---|
| Custom event taxonomy | Strong | N/A (billing events) |
| Funnels and retention | Strong | Cohort revenue charts |
| Entitlements and receipts | Via integration | Strong |
| Paywall and offering tests | Via events | RevenueCat Experiments |
| Warehouse export | Strong | Webhooks + integrations |
For product squads defining dozens of feature events, Mixpanel remains a natural home. RevenueCat remains the answer to "is this user pro?" and "what did we earn this month?"
The pairing breaks down in growth operations, not in raw feature checklists.
Cost one: identity glue
Mixpanel users are often identified by a product account id. RevenueCat customers are keyed by app_user_id. The MMP uses yet another device id. Onboarding may be anonymous for three screens before signup.
Every stack needs glue:
- Alias anonymous Mixpanel distinct id to account id at signup.
- Set RevenueCat
app_user_idto the same stable id. - Forward purchase events to Mixpanel with consistent properties.
- Hope experiment assignment ids appear on both sides.
When glue drifts, funnels lie:
| Symptom | Likely glue failure |
|---|---|
| Trial starts in RC, not in Mixpanel | Client forgot to forward or wrong user id |
| Funnel shows paywall view, RC shows no trial | Different users merged |
| Campaign property missing on purchase | Attribution captured only in MMP |
Engineering can fix each incident. The recurring tax is growth waiting on eng for joins every time they segment "Meta users who started trial but did not open day 2."
A unified analytics profile: acquisition fields, flow steps, and billing milestones on one Rheo identity: does not eliminate the need for careful ids. It removes a whole class of cross-tool exports for subscription funnel questions.
Cost two: renewals without a session
Mixpanel funnels are session- and event-driven. Subscriptions renew on a schedule whether the app opens or not.
| Event | Typical source | Mixpanel visibility |
|---|---|---|
| Trial start | Client + RC | Good if forwarded |
| First paid conversion | Client + RC | Good if forwarded |
| Renewal | Server / RC webhook | Requires pipeline |
| Refund | Server / RC webhook | Requires pipeline |
| Billing retry | Server | Often missing |
Teams bridge this with Segment, Rudderstack, or custom webhooks into Mixpanel. That works. It is another pipeline to monitor, another schema to document, and another place where campaign attribution can drop if renewals are not joined back to first-touch fields.
Rheo Analytics treats subscription lifecycle events as first-class on the same profile that recorded install campaign and onboarding steps: so retention and revenue cohorts by channel do not depend on the user opening the app on renewal day.
Cost three: insight without deployment
Mixpanel answers where users drop off. It does not ship a new onboarding path on Tuesday without:
- Feature flags in the app binary, or
- A release train, or
- A separate remote UI tool
RevenueCat Experiments helps with offering and paywall tests tied to their stack. It does not replace multi-step onboarding graphs, ATT branches, cancellation surveys, or win-back flows.
The operational gap:
Diagnose in Mixpanel → spec in Notion → eng implements → store review → measure againDuring that window, paid traffic still hits the old step four. Convert exists to shorten the loop for journey changes: publish variants OTA, read step-level lift beside revenue on the same profile.
Engage closes the loop on the messaging side: enroll users who dropped on step four into an in-app nudge without exporting cohorts manually.
None of that requires deleting Mixpanel. It requires admitting that product-wide analytics and subscription growth execution are different workflows.
Side-by-side: two operating models
| Question | Mixpanel + RC (+ glue) | Rheo Analytics + Convert + Engage |
|---|---|---|
| Feature adoption funnel | Mixpanel | Mixpanel (or Rheo if instrumented) |
| Onboarding step drop-off by campaign | Join + hope | Native in Analytics |
| Ship onboarding variant | Release or flags | Convert publish |
| Trial start by creative | Forwarded events | Profile + funnel |
| Win-back after cancel | Separate ESP + in-app | Engage + Convert flows |
| Company-wide event taxonomy | Mixpanel standard | May still be Mixpanel |
When to keep Mixpanel + RevenueCat
Stay on the pairing (or add only point integrations) when:
- Analytics platform choice is org-wide and Rheo would duplicate taxonomy governance.
- Your bottleneck is core product features, not onboarding or paywall iteration speed.
- Data team already owns warehouse joins from RC webhooks to campaign dimensions and SLAs are met.
- Experiment surface is RC/Superwall paywalls only, not multi-step journeys.
Mixpanel is genuinely excellent at exploratory analysis and flexible dashboards. RevenueCat is genuinely excellent at subscriptions. Do not rip them out to chase a narrative.
When unified Analytics + Convert + Engage wins
Consider the Rheo growth platform model when:
- Paid UA is material and you routinely decide budget from subscription revenue, not installs.
- Onboarding and lifecycle flows change weekly and App Store latency blocks learning.
- Growth owns segments that must drive both experiments and messaging without weekly CSV exports.
- Step-level funnel metrics must sit next to trial and paid outcomes for the same users.
Read Rheo is a growth platform for the product framing and Product analytics with revenue attribution for attribution specifics.
A pragmatic hybrid
Many teams run:
- Mixpanel for product feature analytics and executive dashboards.
- RevenueCat for billing and entitlements.
- Rheo for subscription funnel analytics, OTA journey experiments, and lifecycle automations on one profile.
Events can be forwarded from Rheo to Mixpanel if you need a single executive funnel view. The goal is not tool count: it is fewer broken joins on the path from install to revenue.
Summary
Mixpanel plus RevenueCat works until identity glue, headless renewals, and slow deployment turn every growth hypothesis into a multi-tool project. Unified analytics on one customer profile does not replace Mixpanel for every question. It replaces the duct tape on subscription growth: campaign → journey → trial → paid → retain.