Google Analytics alternatives, compared

Most people don’t leave Google Analytics over privacy. They leave because they opened it to answer a simple question and couldn’t.

Why people leave GA4

The specific complaints, because “it’s bad” isn’t an argument.

The reports you want are now yours to build. Everything in GA4 is an event, and the analysis most people do lives in Explorations - a canvas you assemble from dimensions and metrics. A real capability if that’s your job. A homework assignment if you publish a newsletter.

Rows disappear without warning. GA4 applies thresholds when counts are low and demographic data is in play. Google, verbatim: “Data thresholds are system defined. You can’t adjust them” (source). The remedy Google offers is widening the date range. The same metric can read differently in two places on the same screen.

Sampling is possible in standard reports. Google’s own comparison of reporting surfaces marks sampling “Possible” for Reports, Insights and Explorations together. Only the BigQuery export is marked “No” (source).

History has a ceiling. Event-level retention on a standard property is 2 or 14 months, and Google says a property that becomes “Large” is automatically reduced to 2 months with older event-level data permanently deleted (source).

Google deleted the last one. Standard Universal Analytics properties stopped processing hits on July 1, 2023, and Google cut off access to current and historical UA data from the week of July 1, 2024 (source). Anyone who didn’t export in time lost a decade of history. That’s the single best reason to care where your data lives.

And the banner. GA4 sets cookies and needs consent in the EU and UK. Everyone who declines is missing from your numbers, and you don’t get to see how many.

One thing we’re not going to tell you: that Google Analytics is illegal in the EU. The 2022 rulings in Austria, France, Italy, Denmark and the Nordics were real, and they were about EU-to-US transfers after Schrems II. The ground under them changed on July 10, 2023, when the European Commission adopted the EU-US Data Privacy Framework, and Google is certified under it. Pages still running that line are stale.

The alternatives, side by side

Annual USD, on annual billing where offered. The “counts” column decides the bill - read it before comparing two prices.

Matomo’s unit is worth repeating: a hit “can be either a tracked page view, an event tracking, a download, an outlink, an onsite search or a content tracking request” - their FAQ. Sites with download and outbound-link tracking often need a tier one or two steps above their pageview count.

Each one, and where it beats us

Matomo

The deepest tool in the list and the oldest - shipping since 2007. Segments, custom dimensions, heatmaps, A/B testing, roll-up reporting, ecommerce, a very large documentation library, and a real on-premise install where the core is free under GPL-3.0. Cloud hosts in Europe.

Where it beats us: feature depth, and the fact that a public-sector procurement office has heard of it. If you need on-premise with an audit trail and a support contract, this is the answer and we’re not close.

The cost: $2,040 a year at a million hits on Cloud, a 46,303-byte gzipped tracker, premium plugins under a proprietary EULA, and a sizing table that goes from 2 GB of RAM at 100,000 hits to 8 GB at a million.

Plausible

The product that made this category mainstream, and the most polished thing in it. Hosted in Falkenstein, Germany, with the company in Estonia. Their tracker measured 1,291 bytes gzipped - smaller than ours. Community Edition is AGPL-3.0-or-later with unlimited sites and seats, and it includes the Stats API.

Where it beats us: maturity and EU hosting, plainly. Years of production, a big customer base, and a straightforward answer for a European buyer who wants data in Europe. We’re a US company and can’t match that.

If Plausible is what you’re actually weighing, seven alternatives to it, compared goes deeper than this entry does.

The cost: the ladder. Custom properties, funnels, ecommerce revenue and the Stats API on Cloud start at Business - $1,390 a year at a million pageviews. Scheduled raw event export is Enterprise-only, so no self-serve plan gets your own raw events out. Community Edition ships twice a year and excludes funnels, ecommerce revenue goals, SSO and the Sites API.

Fathom

Running since 2018, closed source, unusually good at the unglamorous parts. EU isolation is free on every plan: EU visitors’ pageviews route to EU servers where the IP is stripped before anything reaches US infrastructure. Retention is “forever”. 75 sites at the million-pageview tier.

Where it beats us: track record and EU isolation. They’ve been doing this for years, they’ve been audited by buyers with real requirements, and the EU routing is a thoughtful piece of engineering that we don’t have.

The cost: $600 a year at a million, no plan below 100,000 pageviews (so a small blog pays $150), and the product is closed - “Fathom Lite” on GitHub is MIT but has had no code change since January 2023.

Umami

Free, MIT since 2020, self-hostable in two containers, and $20 a month on Cloud Pro with no annual discount - which stays $20 all the way to a million events.

Where it beats us: the license. MIT is more permissive than our AGPL, full stop. If you want to embed, fork or resell without the network-source obligation, Umami lets you and we don’t.

The cost: the unit counts each stored event property, so property-heavy sites multiply fast. And their FAQ says Cloud has email reports and a streaming API the self-hosted build doesn’t.

Pirsch

$540 a year at a million on Standard, unlimited seats, unlimited retention, 50 sites, German company.

Where it beats us: unlimited retention and a German vendor at a low price.

The cost: Standard has no funnels and no teams - those start on Plus at $1,590 at a million. Product is closed; the Go tracking library is AGPL.

Simple Analytics

Dutch, EU-hosted, one of the tidier dashboards in the category. $600 a year at a million.

Where it beats us: EU hosting and a calm interface.

The cost: the datapoint unit counts pageviews, events and each stored event property, and extra seats are $240 a year each.

GoatCounter

A one-person project, EUPL-1.2, hosted free “for reasonable public usage” and funded by donations. Philosophically the closest thing here to us.

Where it beats us: it’s free, and its GDPR page is the most careful writing on the subject in this whole category - including “I am not a lawyer.” Ours is at GDPR-friendly analytics, hedged for the same reason.

The cost: it’s explicitly not for millions of pageviews a day, and the feature set is small.

Cloudflare Web Analytics

Free, cookieless, trivial to switch on if you’re already behind Cloudflare.

Where it beats us: the price, and zero setup.

The cost: shallow reporting, no real goals or custom events, no export, and no way to take your history anywhere.

Feasible - us

The flag we promised. $9.99 a month or $99 a year, one plan: unlimited sites, unlimited team members, unlimited custom properties, and a million pageviews and custom events a month between them, kept five years. Cookieless, AGPL-3.0-or-later, one Go binary on SQLite that runs on 512 MB of RAM. The API, funnels, custom properties, webhooks, the MCP server and raw event export are in every plan and every build - the API package has no plan check in it at all.

Where we lose: no EU hosting, no SSO or SAML, no mobile SDKs, no session replay, no heatmaps, no A/B testing, no cohort or retention report. Our script is 3,377 bytes gzipped against Plausible’s 1,291. And we launched in 2026, so we have no track record - which is a real risk and you should price it in.

Free Google Analytics alternatives

There are several, and they’re free for different reasons. Cloudflare Web Analytics is free and cookieless, and shallow. GoatCounter’s hosted tier is free for reasonable public use, funded by donations. Umami, Matomo and Feasible are all free if you run them on your own server - which is a server, backups, upgrades and a pager, so it’s free the way a puppy is free.

GA4 itself is the other kind of free: your visitors’ behavior goes to an advertising company, and you pay in a consent banner.

Which to pick, by situation

A personal blog. GoatCounter or Cloudflare, free. If you’d rather pay for support and want goals, $99 a year buys you every feature we have.

You spend real money on Google Ads. Stay on GA4 for the ad integration, and if the interface is the problem, run something readable alongside it. Nothing in this list replaces the Google Ads connection.

An EU company with an EU data requirement. Plausible, Matomo Cloud, Simple Analytics or Fathom’s EU isolation. Not us, unless you self-host in the EU.

A freelancer or agency with a lot of client sites. This is where per-site pricing bites hardest. Fathom includes 75 sites; Pirsch 50 with unlimited seats; Feasible unlimited on both. Count your sites before you read the headline price.

A SaaS that needs funnels and custom properties. Check which plan they’re on. Funnels and custom properties are Business-tier on Plausible ($1,390 at a million), Plus on Pirsch ($1,590), and included at $99 here.

A public-sector or heavily regulated buyer. Matomo. Depth, on-premise, EU hosting and two decades of institutional presence.

You want MIT. Umami.

Moving off GA4

The blocker people name most often isn’t price - it’s “I’d lose all my history.”

You wouldn’t, mostly. Export your GA4 reports as CSV and upload the folder, and pages, sources, locations, devices and browsers land on the same charts as your live traffic. A direct Google connection is on the way; today it’s an upload. Universal Analytics history is a different matter: Google cut off access from the week of July 1, 2024, so if you didn’t export then, it’s gone.

The whole process - install alongside GA, run both for two weeks, upload the CSVs, verify, then remove the old tag - is written out step by step in moving off Google Analytics.

Prices move. If something here is out of date, tell us and we’ll fix it. Also worth reading: the wider tool roundup, our head-to-head with Google Analytics, Plausible alternatives, and what Feasible costs.