I've been having problems when analyzing onboarding funnels in reports due to the delay in ID merges
It may take up to 24 hours for this mapping to propagate to all other parts of the system. This means that, in some cases, when analyzing a funnel that spans pre-login and post-login behavior in real-time, some may be shown as dropped-off, even though they’ve performed the conversion event.
I wonder if there's a way to flush the merge immidiately or if someone figured out a way to keep the onboarding funnel work properly realtime
Hi YJ K., 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
There isn’t a way to “flush” or force the ID merge to propagate faster than the documented delay.
For both Original and Simplified ID Merge, Mixpanel updates the single-user Activity Feed in near real time (under a minute), but it can take up to 24 hours for the merged identity to be reflected consistently across all reports, including Funnels.[Simplified FAQ][Original FAQ] This is exactly why you’re seeing users appear as dropped off in onboarding funnels that span anonymous → logged‑in behavior shortly after they convert.[Funnels FAQ]
Given the docs, there’s no supported mechanism to:
trigger an immediate global recomputation of merges, or
bypass the 24‑hour propagation window for funnel calculations.
Workarounds mentioned in the docs are more about implementation hygiene than true real‑time funnel accuracy:
Make sure you call .identify() at registration/login and when reopening in a logged‑in state, and .reset() on logout, so that merges are correct and complete.[Simplified best practices]
Track your own stable user ID as an event and profile property to help debug and verify that the right events belong to the same person, even if the funnel hasn’t fully “caught up” yet.[Simplified best practices]
But the knowledge sources do not describe any official pattern that makes a cross‑anon→auth onboarding funnel fully reliable in real time; you may need to:
accept that those funnels are accurate only after the merge window (e.g. look at “yesterday and earlier”), or
temporarily analyze pre‑login and post‑login steps in separate funnels/Insights reports for real‑time monitoring, then rely on the full cross‑journey funnel for back‑looking analysis once merges have propagated.
If you’d like, describe your exact onboarding flow (events and where identify/reset are called), and I can help you sanity‑check that implementation against the documented best practices.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
