What should I use instead of Google Analytics?

Which job are you actually replacing?

"What should I use instead of Google Analytics" is a hard question to answer well, and the reason is that Google Analytics is not one product. It is four jobs sold under a single tag, and almost nobody needs all four.

Traffic reporting. How many people came, from where, to which pages, up or down on last month. This is what most people look at, most of the time.

Ad attribution. GA4 is plumbed directly into Google Ads. Conversions and audiences move between them, and that plumbing is most of why GA4 is installed on sites whose owners never open it.

Custom event analysis. GA4's data model is events with arbitrary parameters, so with enough setup effort you can answer detailed product questions. Who used the filter, who reached step three, what they did next.

And the fourth job, which GA4 has never done. Telling you which of those visitors was worth your time. Not how many sessions, but which people showed real intent and which were noise.

Work out which of the four you need and the answer stops being a shootout between logos. If it is the first, the privacy-first tools beat GA4 on nearly every axis you care about, and this article will tell you to go and buy one. If it is the second, you are keeping GA4 whether you enjoy it or not. If it is the third, you keep GA4 or move up to a product analytics tool. And if what made you type the search was the fourth, none of the alternatives fix it, because none of them changes what is being counted.

Cost is almost never the reason, incidentally. The standard GA4 tier is free and there is no signal that changes. People leave because the interface is meaningfully harder to use than the one it replaced, because privacy and data residency turned into real requirements rather than nice-to-haves, and because after all the setup the reports still do not say whether any of that traffic mattered.

The privacy-first replacements: Plausible, Fathom, Matomo

These are good products. I am not going to damn them with faint praise on the way to a pitch, because for the first job on that list they are simply the right answer.

Plausible. Open source, EU-hosted cloud service, and built around a single dashboard that fits on screen. The script is tiny, which matters more than it sounds: analytics is often the heaviest third-party thing on a marketing site, and the page-speed difference is measurable rather than theoretical. It supports goals and custom events, so it is not purely a pageview counter, but the design intent is obvious. Read the whole thing in ten seconds and get on with your day.

Fathom. The same shape of product with a different temperament. Hosted, deliberately simple, one screen, goals and events, no consent-banner theatre. Both Plausible and Fathom market themselves as usable without a cookie banner, and for a lot of small sites that is the actual reason for the switch rather than any principled position on privacy. Choosing between them is mostly taste. Run both for a week and keep the dashboard you enjoy opening.

Matomo. The different animal of the three. It has a free self-hosted edition and a paid cloud edition, and it is the closest thing here to a like-for-like GA replacement in breadth: goals, funnels, segments, and a plugin ecosystem for the heavier features. Self-hosting is the whole point for many of its users, because the visitor data then sits on infrastructure you control, which is the only version of a data-residency answer some legal teams will accept.

Be honest about what self-hosting costs, though. You own a database that grows forever, you own the upgrades, and you own whatever happens when it falls over on a Sunday. Pick Matomo when control is a requirement rather than a preference.

What all three deliberately leave out is the second job. None replaces GA4's link into Google Ads. You can run one alongside GA4, the scripts do not conflict, and plenty of teams do exactly that: GA4 stays installed and unopened for the ad platforms, and the daily reading happens somewhere pleasant.

The comparison table

Read that table looking for where each column wins, because every one wins somewhere real. GA4 is free and nothing else here touches its breadth or its ad plumbing. Plausible and Fathom are the nicest daily-use tools on the list. Matomo is the only one that will put the raw data on a machine you own. And exactly one row has a single yes in it, which is the row about the fourth job.

What none of them replace

Every column in that table except the last one counts visits. They count them more privately, more cheaply, or with less pain than GA4 does, and those are genuine improvements worth switching for. What none of them changes is the thing being counted.

The unit of analysis stays the event. A pageview, a session, a goal completion. All of it describes what happened on the site in aggregate, and none of it describes a person. A bounce rate of 62% is a property of a pile of sessions, not a fact about anybody. The visitor who read your pricing page three times this week and the visitor who landed on a blog post and left in four seconds each contribute one session, and once counted they are indistinguishable.

That is what the diagram shows. On the left, what a traffic tool produces: sessions, pageviews, average duration, bounce rate. Every one accurate, none of them answering who wanted to buy. On the right, a verdict per visitor: converted after reading pricing twice, strong intent on a third visit, browsing with no buying signal, left instantly. Same raw behaviour, different unit, different question answered.

You can segment your way partway there, and good analysts do it every day. But every filter leaves you holding a smaller pile of events, never a read on an individual. There is no column for intent in any of these tools because nobody ever measured it, which is the entire subject of AI analytics vs web analytics.

So if you have been switching analytics tools every eighteen months hoping the reports will finally say something about your buyers, the tool was never the problem. Switch for speed, privacy, cost or sanity. Just do not expect a different counter to reach a question counting cannot.

A different answer: classify the visitors instead

The alternative is to stop counting events and read visitors one at a time. That is the approach TribeMap takes, and it is worth saying immediately what it is not: a drop-in GA4 replacement. It does not do attribution reporting for your ad platforms, and it is not the tool to reach for when you need to know exactly how many people saw a page last March.

What it does instead is classify every visit into one of five tribes. Joiner, who converted through a purchase, signup or other key action. Evaluator, who showed strong intent through repeat visits and depth into the funnel. Explorer, engaged across several pages with no conversion signal. Browser, a light browse with fewer pages and lower depth. Bouncer, who left immediately with no engagement. Every classification records a confidence score and the reasoning behind it, so a verdict can be inspected and disagreed with rather than taken on faith.

On the map those five collapse into three colours. Green covers joiner and evaluator, yellow covers explorer and browser, grey is bouncer. The collapse is for reading speed, and the full five-way split is one click away in the detail view. Worth saying plainly, because it is the easiest thing to overclaim: green is not the same as buyers. It includes evaluators who have bought nothing and may never. Green means the behaviour is consistent with someone seriously considering what you sell.

Those classified visitors sit on a living map made of six territories, which are the six ways people arrive: Search, LinkedIn, Email, Social, Direct and Referral. You set the time range you care about, and today updates in realtime. The reason to arrange verdicts as territory rather than as a table is what it makes obvious without being asked. A large territory sending mostly grey is expensive noise however good it looks in a traffic report, and a small territory that is mostly green is worth defending. In a conventional dashboard that comparison is several exports away.

The numbers have not gone anywhere. They live in the ops terminal, for the times when a count really is the right answer.

Underneath all of it the job is closing the gap on three things: WHO converts, which is product and market fit; WHERE they are reached, which is channel; and WHAT moves them, which is message. Installation is one script tag. No personal information is collected and input values are never stored, so what you get is a judgement about behaviour rather than a file on a named human being. Pricing is $49 a month for around 10,000 visitors a month, with a 7-day free trial.

And the honest limit, which applies to everything in this category: a verdict is a judgement under uncertainty, not a fact. A model reading behavioural signals will sometimes call a serious buyer a browser. Treat one verdict as a signal worth checking and the pattern across hundreds as evidence worth acting on. The longer version of that argument, including the tools that beat TribeMap when your buyers are already identified, is in the best AI analytics tools.

So what should you actually do?

Pick by which job you are replacing, not by which logo you liked.

You have a small site and you want simplicity. Buy Plausible or Fathom. They are better at privacy-friendly traffic reporting than anything else described here, they cost less, they load faster, and you will be happier. Do not buy an interpretation engine to find out how many people read your blog.

You need EU data residency or genuine control of the data. Matomo, self-hosted, and accept that you have taken on a server. If the requirement is softer than that, the cloud edition or Plausible's EU hosting will probably satisfy it with less work.

You spend real money on ads. Keep GA4. Nothing here replaces its coupling to Google Ads, and swapping it out to feel better about your stack is a good way to lose money quietly. What is worth adding is something that reads intent, because a channel that looks efficient on cost per click and sends visitors who never engage is the failure your ad reporting is blind to.

You need deep product-event analysis. Keep GA4, or move up to a product analytics tool built for it. The privacy-first tools are deliberately shallower here, and that is a feature rather than an oversight.

You are trying to work out who your buyer actually is. That is the fourth job, and no traffic counter does it at any price. You need something that reads each visitor's behaviour and reaches a conclusion. That is classification, and it is a different purchase from everything above.

Whatever else you do, keep a traffic tool. It is cheap, it will still be right when a model's verdict is wrong, and the two answer different questions. Almost every case of buyer's remorse in this category comes from expecting one of those answers and paying for the other.