Matomo vs PostHog: two privacy-friendly tools, two different jobs
Both are open source and GDPR-friendlier than GA4, which is where the similarity ends. One replaces web analytics, the other explains product behavior.
Matomo and PostHog solve different problems: Matomo is a privacy-first web analytics replacement for GA4, strongest on ownership and compliance; PostHog is a product analytics platform with session replay, feature flags, and experiments. Choose Matomo for marketing-site measurement under GDPR, PostHog for understanding user behavior inside a SaaS product.
- tool1
- Matomo
- tool2
- PostHog
The comparison people actually mean
Matomo and PostHog end up in the same shortlist for one reason: both are open source, both offer EU hosting, and both let you leave the Google ecosystem. From there the products diverge so hard that the real question is rarely "which one" but "which problem do I actually have."
Matomo is web analytics: traffic, sources, campaigns, e-commerce funnels, the GA4 job done with full data ownership. PostHog is product analytics: events inside your app, retention cohorts, session replay, feature flags, experiments. The overlap is thinner than either marketing site admits.
What Matomo is actually good at
Matomo is the mature "GA4, but yours" option. Self-hosted or on Matomo Cloud, the data sits in your database, under your jurisdiction, with no data sent to an ad giant. Reporting covers the classic web questions: acquisition, behavior, conversions, e-commerce, with raw data access and no sampling at standard scales.
Two properties matter for EU businesses specifically. First, ownership simplifies GDPR conversations structurally: no third-country transfer question when the data never leaves your server. Second, configured cookieless, Matomo can qualify for analytics consent exemptions in some jurisdictions (France's CNIL exemption is the known example), subject to configuration rules that your lawyer, not your analyst, should confirm. The full GDPR-angled comparison against GA4 is in GA4 vs Matomo for EU e-commerce.
The honest costs: the interface feels a generation older than PostHog's, some conveniences (heatmaps, funnels on-premise) are paid plugins, and self-hosting is real infrastructure work, not a checkbox.
What PostHog is actually good at
PostHog is built around the question web analytics cannot answer: what do identified users do inside the product over time? Event autocapture plus custom events, funnels, retention, cohorts, session replay to watch the friction, feature flags and experiments to act on it, all in one platform. For a SaaS team it collapses three or four tool subscriptions into one.
It also ships a web analytics module, and it is fine, in the way a multitool's screwdriver is fine. If your primary need is marketing-site reporting for a store, PostHog is the wrong-shaped tool carried by a good pitch.
The honest costs: usage-based pricing needs watching once event volumes grow (replay especially), and the platform's breadth means a team can drown in capabilities it never operationalizes. Buying PostHog is easy; becoming a team that runs experiments is the actual work.
The decision, stated plainly
- E-commerce or content site leaving GA4 for privacy reasons: Matomo. The job is web analytics; buy the tool shaped like the job.
- SaaS trying to understand activation, retention, feature usage: PostHog. Web analytics alone cannot see your revenue mechanics; the reasoning is in web analytics vs product analytics.
- SaaS that also needs marketing-site measurement: PostHog first, and let its web module cover the site until proven insufficient. One tool with one user model beats two tools you have to reconcile.
- Hard data-residency requirements with in-house infrastructure: Matomo self-hosted is the strongest ownership story in mainstream analytics, with the operational bill that implies.
The mistake to avoid
Do not run both to "get the best of each" unless someone owns the reconciliation. Two tools measuring overlapping surfaces with different session models produce two versions of every number, and unowned discrepancies quietly corrode trust in both. If you do run both, split them cleanly: Matomo owns the anonymous marketing site, PostHog owns the logged-in product, and the signup event is the documented border between their territories.
Whichever way you lean, the tool migration is the small half of the project. The events, definitions, and identity handling you bring to it decide whether the new tool reports truth or just new-looking noise, and that part is portable: get it right once and either tool will serve it.