PostHog
All-in-one product analytics, session replay, feature flags, A/B testing.
What it replaces
What PostHog actually is
PostHog is an all-in-one product analytics platform: product analytics, session replay, feature flags, and A/B testing shipped as one open-source package you can self-host or run on the official cloud. Its positioning is aimed straight at Mixpanel and Amplitude, with Google Analytics as the adjacent escape route for teams whose "analytics" is really product instrumentation. If Umami is the minimalist counter and Matomo is the GA replacement, PostHog is the one asking why your analytics, replay, and flags live in three vendors at all.
Live GitHub stats
The PostHog/posthog repository has 39,973 stars, 3,430 forks, and 5,820 open issues. It is primarily Python, was last pushed on 2026-09-27, and the license situation deserves a pause: GitHub reports NOASSERTION because the repo mixes licenses — the project describes its licensing as MIT plus AGPL — so run your compliance check against the actual file headers rather than a one-line label. Those numbers describe one of the most active analytics codebases in this catalog; the issue count is the price of that surface area.
What it's good at
PostHog is for product teams, not site owners. The center of gravity is instrumenting a product: funnels, retention, replay, experiments, flags. If your requirement is "how many people visited my marketing site and from where," PostHog will answer that too, but buying a platform to count pageviews is bringing a data center to a picnic. The official cloud has a free entry tier (listed at $0), which makes it unusually easy to evaluate before committing to anything.
The comparison that matters most is against itself-as-cloud: self-hosting PostHog means owning a genuinely non-trivial deployment in exchange for data control. Teams that just want dashboards should stop at the cloud tier or pick one of the lighter tools in this catalog.
What to watch for
Deploy difficulty is rated 4 out of 5, the heavy end here — self-hosting PostHog is a multi-service project, not a single container you can forget about. The 5,820 open issues are the flip side of a fast-moving platform: lots of surface, lots of edge cases, lots of GitHub notifications. And the MIT + AGPL mix means "it's open source" needs a follow-up question in any procurement conversation. None of this is disqualifying; all of it is why PostHog is a platform decision rather than a script tag.
How to deploy
There are two supported paths: the official cloud (free tier listed at $0) and self-hosting for teams that need data control. Self-hosting is the 4-out-of-5-difficulty option — budget real operational time for it and follow PostHog's own self-host documentation rather than a forum recipe. There is no maintained deployment guide on this site yet, so start from the official docs.
Related tools
The rest of the analytics lane: Umami for the lightweight counter, Plausible Analytics for privacy-first web analytics, and Matomo for the full GA-replacement platform. Head-to-heads: Matomo vs PostHog, Plausible vs PostHog, PostHog vs Umami, plus the pipeline side with PostHog vs RudderStack and Jitsu vs PostHog. For the SaaS-lens view, see Google Analytics alternatives, Mixpanel alternatives, and Amplitude alternatives.