PostHog vs Amplitude vs Mixpanel: Product Analytics Compared (2026)

I keep seeing "PostHog vs Amplitude vs Mixpanel" framed as a features checklist. Twenty rows of green checkmarks, a vague recommendation to "evaluate based on your needs," and a call to action. That's not useful. These three tools were built by different people, for different buyers, with different assumptions about who owns the data and who pays for what. The checkmarks overlap, but the philosophies don't.

I've run all three in production — sometimes two at once on the same event stream. Here's what I actually learned, including the pricing math nobody wants to publish.

Quick decision matrix

Before the deep dive, the short version:

• Pick PostHog if your engineers drive product decisions and you want analytics, session replay, feature flags, and experiments in one open-source tool you can self-host.
• Pick Amplitude if your product team includes non-technical PMs who need deep behavioral analysis, predictive analytics, and a built-in CDP to sync audiences to your marketing stack.
• Pick Mixpanel if you want clean, fast event analytics without paying for features you won't use, especially if you're a B2B company that needs account-level analysis.
• Pick none of them if your real question is "which acquisition channel produces the highest-LTV customers" — that requires joining product events with revenue and CRM data, which is a different problem entirely.

Architecture and data ownership

These three tools make fundamentally different bets about where your data should live.

PostHog is open-source (MIT-licensed core, with some enterprise features under a proprietary license). The analytics engine runs on ClickHouse — a columnar database built for fast aggregation over billions of rows. You can self-host the entire stack on Kubernetes: ClickHouse, Kafka, PostgreSQL, Redis, the app servers. Your events never leave your network. For healthcare, fintech, or any company with a strict GDPR interpretation, that's not a nice-to-have — it's a compliance requirement. The GitHub repo has over 22,000 stars, 50+ active contributors, and ships weekly releases. PostHog Cloud exists too, and most startups use it — running a production ClickHouse cluster is real operational work.

Amplitude is closed-source and cloud-only. No self-hosting option. Your events go to Amplitude's infrastructure, period. They've built a warehouse-native mode that lets you run queries against your own Snowflake or BigQuery, but the application layer still lives in their cloud. Amplitude is a public company (AMPL) with deep enterprise roots — SSO, audit logs, role-based access, SOC 2 compliance baked into the vendor relationship.

Mixpanel is also closed-source and cloud-only. They recently rebuilt their query engine on Google BigQuery, which made their backend faster and cheaper to operate. Mixpanel was acquired by ServiceNow in late 2024, which changes the long-term calculus — you're now betting on ServiceNow's continued investment in a product analytics tool that competes with ServiceNow's own analytics capabilities. Whether that's reassuring or concerning depends on your risk tolerance.

The data ownership spectrum is clear: PostHog gives you full control (self-host with your own infrastructure), Amplitude and Mixpanel hold your data in their clouds. If data sovereignty matters to your business, PostHog is the only real option here.

Event tracking: how it actually works

All three tools follow the same core model: you instrument your app to fire named events with properties, and the platform ingests, stores, and lets you query them. The SDKs are similar enough that switching between them is a weekend of find-and-replace, not a rewrite.

Here's what tracking a checkout event looks like in each:

PostHog:

import posthog from 'posthog-js';
posthog.init('phc_YOUR_KEY', { api_host: 'https://app.posthog.com' });
 
posthog.capture('checkout_completed', {
  plan: 'pro',
  amount: 49.99,
  currency: 'USD',
});

Amplitude:

import * as amplitude from '@amplitude/analytics-browser';
amplitude.init('YOUR_API_KEY');
 
amplitude.track('Checkout Completed', {
  plan: 'pro',
  amount: 49.99,
  currency: 'USD',
});

Mixpanel:

import mixpanel from 'mixpanel-browser';
mixpanel.init('YOUR_TOKEN');
 
mixpanel.track('Checkout Completed', {
  plan: 'pro',
  amount: 49.99,
  currency: 'USD',
});

The syntax is nearly identical. All three support JavaScript, React Native, iOS, Android, Python, and Flutter. PostHog adds Ruby, Go, and Node server-side SDKs. Mixpanel covers Ruby, Java, and Unity. Amplitude has Java and Go. For most web and mobile apps, SDK coverage isn't a differentiator — all three have you covered.

The real differences show up in what happens after the event is captured. PostHog stores everything in ClickHouse and exposes it via HogQL (their SQL dialect), so you can write raw queries against your event data. Amplitude routes events through its behavioral graph — an internal data structure that pre-computes relationships between user actions, which powers its more advanced analytical features. Mixpanel sends events to its BigQuery-backed engine, optimized for fast segmentation and funnel queries.

Analytics capabilities

The core analysis primitives — funnels, retention, trends, user paths — exist in all three. The differences are in depth and polish.

Funnels: Mixpanel's funnel builder handles the most edge cases. You get ordered and unordered steps, per-step conversion windows, and exclusion events (show me users who did A then B but did NOT do C between them). Amplitude's funnels are similarly powerful, with the added benefit of behavioral cohorts — you can build a funnel that only includes users matching a complex behavioral definition. PostHog's funnels are functional and improving but less refined at the edges.

Retention: All three offer N-day, unbounded, and custom retention. Amplitude's lifecycle analysis goes a step further — it categorizes users into new, current, resurrected, and dormant segments automatically, which gives you a clearer picture of whether growth is healthy (new users) or sugar-coated (resurrected users returning briefly). Mixpanel has a similar feature but Amplitude's visualization is cleaner.

Behavioral cohorts and predictions: This is where Amplitude genuinely leads. The behavioral graph powers queries like "users who did X at least 3 times in the last 7 days but never did Y" — with a drag-and-drop interface that PMs can use without SQL. Amplitude also has predictive analytics: churn likelihood scores, conversion probability, anomaly detection. Neither PostHog nor Mixpanel offers prediction out of the box.

Group analytics (B2B): If you're building a B2B product, Mixpanel's Group Analytics is the most mature. You define entities — companies, workspaces, teams — and every report type works at the group level. "How many accounts activated" instead of "how many users clicked." PostHog supports groups, and Amplitude handles it through its Accounts add-on, but Mixpanel has been iterating on this the longest.

Session replay: Only PostHog has it. Amplitude and Mixpanel require a separate tool (FullStory, Hotjar, LogRocket). PostHog's killer workflow is going from a funnel drop-off directly into session recordings of the users who dropped off — two clicks from "40% bail at step 3" to watching exactly what went wrong.

Pricing reality check

Here's the math nobody publishes. Pricing as of mid-2026:

At 5M events per month (a typical Series A startup):

• PostHog Cloud: 1M events free, 4M at ~$0.00031/event = ~$1,240/month. Add session replay and feature flags and you're looking at $1,500-2,000/month total depending on usage.
• Amplitude: If you're under 50K monthly tracked users, the free Starter plan covers you. If you're over, Growth starts at ~$49/month but scales with MTU count — realistically $500-2,000/month at this volume depending on the negotiated rate.
• Mixpanel: Free. The 20M events/month free tier covers 5M with room to spare. You pay nothing until you quadruple your volume.

At 50M events per month (a scaled product):

• PostHog Cloud: Roughly $10,000-13,000/month. PostHog's per-event tiers decrease at higher volumes, but the bill is still significant.
• Amplitude: Enterprise contract territory. Expect $40K-80K/year depending on MTU count and which products you license (Analytics only vs. Analytics + Experiment + CDP).
• Mixpanel: 30M events over the free tier. Growth pricing kicks in — estimate $2,000-5,000/month depending on your specific plan and add-ons.

The counterintuitive takeaway: Mixpanel's free tier is the most generous for raw event volume. PostHog's free tier is the most generous if you also need session replay, feature flags, and experiments bundled in. Amplitude's free tier is the most generous in MTU terms (50K tracked users), but the jump to paid is steep.

But comparing analytics-only pricing misses the real cost equation. If you use PostHog for analytics + session replay + feature flags + experiments, you're replacing four separate vendor contracts. A stack of Mixpanel ($0) + FullStory ($10K/year) + LaunchDarkly ($10K/year) + Statsig ($8K/year) costs more than PostHog's all-in-one bill, and the integration between separate tools is always worse than having everything in one database.

All-in-one vs. best-of-breed

This is the philosophical divide. PostHog 3.0 took the all-in-one bet even further by adding a data warehouse feature — you can query Stripe, HubSpot, and Postgres data alongside your product events directly in PostHog. They're trying to be the single pane of glass for product teams. Amplitude is building a platform too (Analytics + Experiment + CDP), but each product is priced and sold separately. Mixpanel stays focused: event analytics, done well, nothing more.

The trade-off is real. PostHog's session replay isn't as refined as FullStory. Its feature flags aren't as battle-tested as LaunchDarkly at Fortune 500 scale. Its analytics isn't as polished as Mixpanel's for pure funnel work. But the integration between components — watching a session recording for a user in a specific experiment variant who dropped off at a specific funnel step — is genuinely better than anything you can build by stitching four separate vendors together.

If you're the kind of team that picks one vendor and goes deep, PostHog. If you're the kind of team that picks the best tool for each job and wires them together, Mixpanel for analytics and your preferred tools for everything else.

Three-way comparison table

Who should pick what

I'll be specific.

Early-stage startup (under 30K MAU, small engineering team): Start with Mixpanel. The 20M free events cover you for a long time, the UI is fast and intuitive, and you won't spend weeks configuring features you don't need yet. If you grow into needing session replay or feature flags, evaluate PostHog then.

Engineering-heavy team that ships fast: PostHog. Your engineers will appreciate HogQL, the open-source codebase they can inspect, and having feature flags integrated with analytics so experiment results show up automatically. The consolidation of four tools into one saves real integration work.

B2B SaaS with an account-based motion: Mixpanel, specifically for Group Analytics. When your sales team asks "which accounts are engaged" and your board asks "how many companies activated this quarter," Mixpanel's account-level analysis handles this cleanly.

Growth team with dedicated PMs and analysts: Amplitude. Behavioral cohorts, predictive churn scores, lifecycle analysis, and the CDP for syncing audiences to ad platforms and email tools — these features exist because Amplitude was built for exactly this team.

Regulated industry (healthcare, fintech, EU data residency): PostHog, self-hosted. It's the only option where events never leave your infrastructure. The Kubernetes deployment is operational work, but it eliminates an entire category of compliance conversations.

Budget-constrained, analytics-only: Mixpanel at low volume (free tier is massive), PostHog if you need more than just analytics.

The gap all three share

All three tools tell you what users do inside your product. None of them are built to connect that behavior to what happens outside it — revenue collected in Stripe, pipeline in your CRM, the ad spend that drove the signup in the first place. You can see that 500 users completed onboarding. But how many became paying customers? Did the cohort from your latest campaign retain better than organic? Those questions require joining product events with billing and acquisition data across systems.

Fastero connects product analytics data with revenue and operational data from Postgres, Stripe, HubSpot, and ad platforms — so you can ask questions that span tools rather than staying inside one event stream.

For deeper two-way comparisons, see our head-to-head posts: PostHog vs Mixpanel, PostHog vs Amplitude, Mixpanel vs Amplitude, and Heap vs Mixpanel.

Try Fastero free — connect your product database and ask questions in plain English — get the analytics you need without adopting another platform. No credit card required.

Last updated: July 2026.