PostHog

What is PostHog?

PostHog is an open source product analytics platform that captures every event in your app and turns it into funnels, retention charts, and dashboards. It bundles session replay, feature flags, A/B testing, surveys, and error tracking into one tool, so you get the whole product-intelligence stack instead of stitching together separate SaaS products.

What is PostHog best for?

Product and engineering teams that want to understand user behavior and ship features safely from a single platform. It fits builders who’d otherwise pay for Mixpanel, LaunchDarkly, and a session-replay tool separately — and teams that want their analytics data kept in an instance they control rather than scattered across vendors.

What can PostHog do?

• Event-based product analytics with funnels, retention, paths, and a SQL query editor
• Session replay to watch real user sessions and debug UX issues
• Feature flags for gradual rollouts, targeting, and kill switches
• Experiments (A/B testing) with built-in statistical significance
• Surveys, error tracking, and a GA-style web analytics dashboard
• A data warehouse that syncs external sources like Stripe and HubSpot, plus pipelines to export events to 25+ tools
• SDKs for most languages and frameworks, plus autocapture so you can start without instrumenting every event

Is PostHog free?

Yes — PostHog is free to self-host under the MIT license, and PostHog Cloud has a generous free tier that resets monthly (1M analytics events, 5K session recordings, 1M feature flag requests, and more). Beyond the free tier, Cloud is usage-based pay-as-you-go, and each product has its own billing limit so you can’t be charged by surprise.

Where does PostHog fall short?

• Self-hosting is deliberately limited. PostHog steers almost everyone to Cloud: the open source hobby deploy is only meant for roughly 100k events a month, comes with no support or guarantees, and is discouraged for anyone expecting over ~1M events.
• Paid features are Cloud-only. The self-hosted edition gives you free-tier functionality; advanced plan features live on PostHog Cloud and aren’t available in the open source deployment.
• It’s a heavy stack to run yourself. PostHog is built on ClickHouse, Postgres, Kafka, and Redis — far more infrastructure than a lightweight analytics tool, which is a large part of why the project pushes self-hosters toward Cloud.

What does PostHog replace?

PostHog is a self-hosted alternative to Mixpanel, Amplitude, and Heap for product analytics, and to FullStory for session replay. Because it also covers feature flags and experiments, it can stand in for a separate flags product too. If you only need lightweight, privacy-first website stats, a simpler web analytics tool like Plausible, Umami, or Matomo is a lighter fit.

FAQ

Is PostHog open source? Yes. The core is MIT-licensed and the full source is public; some paid features live in a separately-licensed ee directory, and a fully open source fork (posthog-foss) is also available.

Can I self-host PostHog for free? Yes, via the Docker hobby deploy — but it’s officially for small workloads (around 100k events/month) with no support, so most teams past that point are better off on Cloud.

Is PostHog a good Mixpanel or Amplitude alternative? For most teams, yes — it covers the same event analytics, funnels, and retention, and adds session replay and feature flags in the same tool. The main trade-off is that scaling self-hosted PostHog is harder than a hosted plan.

What do I need to run PostHog myself? A Linux host with Docker and about 4GB of memory for the hobby deploy; under the hood it runs ClickHouse, Postgres, Kafka, and Redis.