Best Self-Hosted Alternatives to Google Analytics
Google Analytics is free, everywhere, and the reason most sites have a cookie banner. What sends people looking elsewhere is rarely the price: it is the consent notice, the sampling and data retention limits on the free tier, a dashboard that has to be learned rather than read, and the fact that the visitor data goes to Google before it comes to you. The self-hosted tools below all answer the same question — how is my site doing — on a box you own, and most of them do it without cookies at all.
Google Analytics is free — but your data lives on its servers, not yours. These researched, open-source alternatives give you the same workflow on hardware you own — here's what each needs to run.
- The easiest Google Analytics swap
- Umami — one script tag, one clean dashboard with custom events, MIT-licensed Next.js on PostgreSQL in 1 GB — the pick for most sites.
- The like-for-like replacement
- Matomo — goals, e-commerce tracking, segments and raw visitor data you own, in a PHP + MariaDB app on 2 GB — funnels and heatmaps are paid plugins.
- GA's report plus sessions, funnels and replays
- Rybbit — an AGPL-3.0 stack installed by one script (Caddy included) on the docs' 2 GB floor — needs a domain, because the tracking script must be served over HTTPS.
- The proven privacy-first dashboard
- Plausible — the Community Edition of the best-known cookieless analytics tool — Elixir, ClickHouse and PostgreSQL on 2 GB, AGPL-3.0.
- Analytics with performance and error tracking built in
- Swetrix — a cookieless AGPL-3.0 Community Edition in a project-maintained compose file on 2 GB — alerts, email reports and revenue/experiments/AI stay cloud-only.
- The smallest thing that counts
- GoatCounter — a single Go binary or container with SQLite, a no-JavaScript tracking pixel and, per its own README, no GDPR notice needed — EUPL-1.2.
- Product analytics, if GA was standing in for Mixpanel
- OpenPanel — funnels, cohorts, user profiles, session replay, A/B tests and mobile SDKs under AGPL-3.0 — the heaviest install here, seven containers behind Caddy.
The alternatives
What you give up by leaving Google Analytics
Leaving Google Analytics means giving up its integrations with Ads and Search Console, its audience modelling, and a service that never needs patching. The alternatives split by weight. Umami, GoatCounter and Plausible are lightweight dashboards that replace the everyday GA report; Matomo is the like-for-like replacement with goals, e-commerce and raw visitor data; Rybbit, Swetrix and OpenPanel add sessions, funnels, replays and product-analytics features GA never had in its free tier. Every one of them is one more server to run, and the ClickHouse-backed ones want 2 GB of RAM before you store a single pageview.
Paid link — we earn a commission if you shop through it.
Head-to-head
A single Go binary on a SQLite file vs. the lightest modern analytics dashboard.
GA-grade depth vs. analytics you can forget you're running.
Polished and opinionated vs. featherweight and deeper.
Common questions
Are these Google Analytics alternatives really free?
The software is open-source and free to run. Your only cost is the server it runs on — often a few dollars a month, less than most SaaS subscriptions.
How much server do I need to self-host one?
Most of these run comfortably on a small VPS — 1–2 GB of RAM is enough for a single-user or small-team setup. Each app's page lists its minimum RAM.
Will I lose features by leaving Google Analytics?
It depends on the app. The alternatives below cover the core workflow; polish and integrations vary, so check each one's stack and difficulty before you commit.