Migrating to Amplitude

Amplitude is a data-graph, not analytics. Teams that migrate from GA4 expecting prettier reports under-use the Spaces taxonomy, skip Govern, and overpay against the $300K MTU cap on Plus. The destination only earns its keep when the org commits to a quarter of taxonomy work, named owners on every event family, and the discipline to retire dirty events instead of stacking new ones on top. Read the source map, then click into the matching pair page — only the GA4 path is currently live.

When Amplitude is the right destination

Amplitude rewards organizations that treat instrumentation as a product surface — versioned, governed, named-owner, retired-on-schedule. The product is built around three tightly coupled ideas: Spaces (workspace + taxonomy boundary), Govern (event-spec ownership and approval workflow), and the data-graph itself (events, identities, behavioral cohorts as first-class entities). Pick Amplitude when those three primitives match how the org already wants to operate, not as a wishlist for how it might operate one day.

• Org-wide self-service mandate, not just a query queue for the analyst team. Amplitude makes sense when product, marketing, growth, lifecycle, and support all need to ship their own charts without a ticket. The whole pricing model — MTU-based, not query-based — assumes 30-200 weekly active users in the dashboard. Below 10 weekly active users you are paying a six-figure annual contract for what one analyst could deliver on Mixpanel free or self-hosted PostHog.
• You need a governance hierarchy: Spaces, taxonomy charter, named owners. Spaces partition the data-graph by product line or business unit. Govern enforces that every event has an owner, a description, a status (live, deprecated, sunset), and an approval trail. This only earns its complexity when the org has 4+ event-emitting teams who would otherwise step on each other's namespaces. One team, one product → use Mixpanel.
• CDP + analytics consolidation considered (drop Segment or RudderStack possible). Amplitude's CDP add-on (formerly mParticle's Cortex stack, acquired 2023) is now a credible Segment replacement for orgs already paying $40-80K/year for Segment Business. The pitch is one event-spec source-of-truth feeding both the warehouse and Amplitude analytics. Whether the consolidation actually saves money depends on Segment-source count and warehouse-destination count — model it before signing.
• You commit to the org-shift: training, charter ownership, taxonomy enforcement. Amplitude payoff lands in months four through nine, not week two. Budget a quarter for the taxonomy charter, a quarter for the rollout, and a quarter for the cleanup pass. Teams that try to "just turn it on like GA4" end up with a $50K/year tool used like a free one and a Slack channel full of complaints about dirty data.

When Amplitude is the wrong destination

Four scenarios where Amplitude is the wrong choice regardless of how good the demo looked. The shape of the wrong-fit is consistent — it shows up six months in, when the contract is signed and the taxonomy is dirty.

• Solo founder, under 50 DAU, or pre-product-market-fit. Amplitude Starter is free up to 10K MTU but the value is governance and self-service across many users. With one user you get a worse Mixpanel. Mixpanel free covers 1M monthly events and ships funnels and retention out of the box; or stay on GA4 free until the team is large enough that governance pays back. Amplitude before 50 DAU is a tool looking for a problem.
• No governance discipline — you will ship dirty taxonomies. If the org has zero appetite for "event-spec review meetings" and event names land in production via "ask the developer who wrote it last," Amplitude amplifies the mess. The data-graph rewards charter discipline and punishes its absence. Mixpanel is more forgiving of a dirty taxonomy; the Lexicon helps but does not require approval workflows. Pick Mixpanel; come back to Amplitude when the org is ready.
• Price-sensitive — the $49 sticker hides the real bill. The Plus tier starts at $49/month with a $300K MTU cap. Above 300K MTU the price jumps to a custom Growth quote that lands between $30K and $90K/year for typical B2B SaaS workloads, and that is before any add-on. Add Experiment and CDP and Session Replay and the all-in is $80K-$150K/year for a mid-sized SaaS. If $50K/year of analytics tooling is not in the budget, do not start the eval.
• Marketing-site-only — no product, no logged-in users. Amplitude's whole identity model assumes a user_id from authentication. For a content site, agency portfolio, or pre-launch landing page, the value disappears: no funnels worth running, no cohorts worth defining. Use Plausible, Fathom, or Rybbit instead. Amplitude on a marketing site is a $50K/year traffic counter.

From which source — pick your starting line

Five common origin points and their dedicated pair pages. Only the GA4 path is currently live; the others are scheduled.

If your source is none of those — Adobe Analytics, Pendo, Snowplow self-host — file an issue on GitHub. The pair page roster is built by request.

Amplitude pricing — the four-tier reality

Pricing is the single most-misread part of an Amplitude eval, because the $49 number on the public pricing page covers maybe 15% of what a typical mid-market deployment actually pays. The honest math, by tier:

Add-ons are separately priced and are not optional for most use cases. Experiment (the A/B testing module) lands $20K-$50K+/year. CDP add-on (the Segment replacement) lands $30K-$80K+/year depending on source count. Session Replay is per-MTU on top of the base contract. A typical mid-market all-in — Growth + Experiment + CDP — sits between $80K and $150K/year. The pricing page does not show the bundle reality; the AE quote does.

The $300K MTU cliff on Plus is the one number to memorize. A B2C app with 200K monthly active users plus website visitors crosses 300K MTU within a quarter of launch. The price change from $49/month Plus to a $40K/year Growth contract is a 70× jump, not a 10× one. Build the financial model assuming you land on Growth within twelve months, not on Plus forever.

Amplitude's three personalities — what are you actually migrating to?

Amplitude is sold as one product but functions as three. Most migrations target Personality 1 (Analytics) and discover Personalities 2 and 3 mid-rollout. Knowing the three up front avoids a year-two re-architecture.

Most teams migrate the Analytics personality, sign a one-year Plus or Growth contract, and only consider CDP and Experiment in year two when the Segment renewal lands or the LaunchDarkly bill grows past $30K. The bundling pitch from the AE will start in month nine; it is a real cost question, not a marketing tactic, and it deserves a real model.

FAQ

Sibling destinations

Other destinations covered on this site. Amplitude fills the deep-governance, multi-team-org slot; the siblings cover different priorities — speed, simplicity, self-host, marketing-site-fit.