Most lists of Google Analytics alternatives are the same eight products described in the same neutral paragraphs, which is useless — the products are genuinely different and the right one depends on facts about you.
Here they are sorted by situation instead.
First: should you leave at all?
Three reasons to stay, stated properly because the honest answer is sometimes "do not bother".
You depend on Google Ads attribution. The integration between GA4 and Google Ads is real and nothing replicates it. If a meaningful share of your revenue comes through Ads and you optimise against that data, leaving costs you something concrete.
You have significant custom implementation. Years of event tracking, audiences and integrations represent real investment. Rebuilding it has a cost that a $9/month saving does not cover.
Nobody looks at it anyway. Blunt, but common. If your analytics is a compliance artefact rather than a decision input, migrating it is work that produces nothing. Turning it off entirely is at least coherent.
If none of those apply, the reasons to leave are usually consent banners, the interface, sampling, or the fourteen-month retention default. All are legitimate.
If you want it to be somebody else's problem
Plausible Cloud and Fathom are the two to consider. Both cookieless, both well designed, both around $9–14 a month to start and rising with traffic. Plausible is open source and self-hostable if you later change your mind; Fathom is not.
Straightforwardly the right answer if you do not want to own any part of running this. The cost is that it never stops — about $228 a year for Plausible, $180 for Fathom, forever — and that funnels, revenue and API access typically sit on the plans above the entry tier, so the advertised price is often not the price. The longer comparison.
If you want it in your own infrastructure
Matomo is the most complete option. It is a genuine GA4 replacement feature-for-feature, with heatmaps, session recording and A/B testing on top. It is also a PHP and MySQL application that needs a real server, real maintenance and real resources, and its cookieless mode is an option rather than the design.
Right if you want everything and have someone to run it.
The Cloudflare Workers projects — Counterscale, Edgemetry, and paid options including this one — are the lightweight end. They run in a cloud account you already have, often inside a free tier, and need no server. They vary enormously in what they can do and in what limits they hit first. Compared individually.
Umami and PostHog sit in between. Umami is a clean, simple self-hosted tool on Postgres. PostHog is a product-analytics platform — session recording, feature flags, experiments — which is a superset of what most people mean by web analytics and a lot to run if you only wanted pageviews.
If your problem is specifically the consent banner
This is the most common trigger and it narrows the field usefully.
Anything that stores nothing on the device sidesteps the ePrivacy trigger entirely. That includes Plausible, Fathom, Cloudflare Web Analytics and the cookieless self-hosted options; it does not automatically include Matomo or Umami, both of which can be configured either way.
The test is whether the tool writes a cookie, uses local storage, or fingerprints. What cookieless has to mean to be worth anything.
If your problem is that you cannot see AI traffic
New in the last two years and increasingly the reason people move.
Every option above measures with a browser script, and therefore cannot see AI crawlers at all. Most also misfile assistant referrals as campaigns, because ChatGPT tags its outbound links.
If this is your problem, the shortlist is short: a tool that classifies AI referrers correctly, and — for crawler visibility — one that measures in your request path, which in practice means self-hosted.
Whatever you choose, take your history
The most common regret is not the tool. It is switching in January and having no comparison for the previous year.
Export from GA4 before you go, and pick something that can import it. This is deeply unglamorous and it is the difference between a switch that costs you nothing and one that costs you your past. How to do it without losing the history.
A short decision path
- Google Ads is central → stay on GA4, add something else alongside if you want the interface.
- Want it handled, happy to pay forever → Plausible or Fathom.
- Want everything, have someone to run a server → Matomo.
- Want it in your cloud account, minimal ops → a Workers-based tool.
- Just want the banner gone → any cookieless option, the cheapest one you like.
- Need AI visibility → self-hosted, in the request path.
Common questions
What is the best Google Analytics alternative?
There is no single answer, because the products differ more than the lists suggest. Plausible or Fathom if you want it hosted and are happy paying indefinitely. Matomo if you want full feature parity and have someone to run a server. A Workers-based tool if you want it in your own cloud account with minimal operations. And if Google Ads attribution is central to your revenue, staying on GA4 is the right call.
Will I lose my data if I move away from Google Analytics?
Only if you leave without exporting it. GA4 lets you export reports as CSV, and some alternatives can import that — daily totals or a breakdown by page, country, device, browser, source or campaign. Do this before you switch. The most common regret after migrating is not the new tool, it is having no year-on-year comparison for the following twelve months.
Do any Google Analytics alternatives show AI traffic?
Partially. Any tool can show referrals from ChatGPT or Perplexity if it classifies those hosts before its campaign rule — most do not by default, because ChatGPT tags its links with a UTM parameter that wins. Showing which AI crawlers read your pages is different: it requires measuring the HTTP request rather than running a script, so only tools running in your own request path can do it.