Webclat / PostHog Practice

Amplitude to PostHog migration: mapping cohorts and keeping trend lines intact

Amplitude's strength is deep behavioral cohorting; the migration risk is losing that analytical depth in translation. The fix is deliberate event and cohort mapping before cutover, not a blanket export-and-hope.

What makes this migration direction specific

Amplitude implementations that have been running for a while tend to accumulate genuinely sophisticated cohort definitions and behavioral segments. The migration risk isn't moving events - it's losing the analytical logic behind those cohorts if they get recreated carelessly in PostHog rather than mapped deliberately.

The mapping work, specifically

Amplitude conceptPostHog equivalentWhat needs deliberate mapping
Events & event propertiesEvents & event propertiesNaming convention often differs; this is the point to fix drift, not just copy it
User propertiesPerson propertiesIdentity resolution logic (how a user is identified across sessions/devices) needs re-verification
Behavioral cohortsCohorts / saved insightsCohort logic must be recreated explicitly - this is the highest-effort mapping step
DashboardsDashboards / insightsRebuilt against the new event schema, verified against Amplitude's numbers during parallel run

The sequence we run

  1. Inventory every Amplitude event, property, cohort, and dashboard actually in active use - not everything that was ever created.
  2. Map each to a PostHog equivalent, applying the autocapture vs. explicit events framework to clean up anything that drifted.
  3. Run both platforms in parallel, reconciling key funnels and retention curves before anyone treats PostHog as authoritative.
  4. Cut over with Amplitude left reachable read-only for historical reference, not switched off.

Where self-hosting fits into this decision

A migration off Amplitude is also a natural point to evaluate self-hosted PostHog, since you're already re-architecting the analytics layer - worth scoping alongside the migration itself rather than as a separate later decision.

Frequently Asked Questions

Will our Amplitude cohorts translate directly to PostHog?

Not automatically - Amplitude's cohort definitions need to be recreated as PostHog cohorts or saved insights, which is a mapping exercise, not an import. This is the step most migrations underestimate.

Can we keep historical Amplitude data visible after migrating?

Amplitude data typically stays accessible on Amplitude itself (read-only) for historical reference, while go-forward tracking moves to PostHog - full historical backfill into PostHog is possible but adds scope worth weighing against the value of unified history.

How do we avoid a gap in trend lines during cutover?

A parallel-run period where both platforms track simultaneously lets you verify PostHog's numbers match Amplitude's before treating PostHog as the source of truth, avoiding a discontinuity in dashboards teams rely on.

Do you run this migration from the Amplitude side too?

Yes - we operate a dedicated Amplitude practice at ampl.webclat.com alongside this PostHog practice, so this specific migration direction is one we work from both platforms, not just the receiving end.

Migrating off Amplitude?

We map your cohorts and events deliberately and run the parallel verification before cutover - from a practice that works both platforms.

Scope My Amplitude Migration