How to choose a Google Analytics 4 alternative

Search for a GA4 alternative and you will find a great many lists of ten tools, most of them written by somebody paid per signup and most of them reaching a conclusion by the third paragraph. This is not one of those. It is the set of questions that actually decide the answer, in the order they should be asked, and the first one may well tell you to stay where you are.

What Google Analytics 4 is genuinely good at

Any comparison that opens by attacking GA4 is selling something. It is free at a scale that would cost real money elsewhere, it exports raw events to BigQuery, and its integration with Google Ads is not matched by anything else in existence. Its ecosystem is enormous. Whatever your stack is, somebody has already connected it.

The reasons people leave are narrow and specific: the consent banner it requires in Europe, sampling and thresholding that make small numbers unreliable, a data model that takes a weekend to learn, and an interface that answers questions you did not ask before it answers the one you did.

Question 1. Do you buy traffic from Google?

If you run Google Ads, the conversation is over. Keep GA4. Conversion imports, audience building and bid optimisation all depend on it, and no privacy-focused alternative replaces that pipeline. What you can do is run a second, simpler tool alongside it for the day-to-day questions. That is a common and entirely sensible setup: GA4 as the advertising instrument, something else as the one you actually open.

Question 2. Are the numbers counted or estimated?

This is the question founders underrate, and at small scale it matters more than privacy does. GA4 applies sampling to explorations above a processing threshold and suppresses rows whose volume is too low to report. Both behaviours are defensible at Google's scale and corrosive at yours. A segment of eleven people is exactly the segment an early-stage founder needs to look at, and it is exactly the one that gets suppressed.

Ask any candidate directly: is every figure computed from recorded events, with no sampling, extrapolation or thresholding? It is a yes or no question, and most privacy-focused tools, DataSpice included, answer yes.

Question 3. Does it store anything on the visitor's device?

This is the testable form of "privacy-friendly". Not *does it set a cookie*, because several tools moved to `localStorage` or to device fingerprinting and kept the marketing language. Storing or reading anything on the device is what triggers the consent requirement, whatever the mechanism is called. The full reasoning is in the cookie banner article. The short version is that a yes here means you still need a banner, and the banner is usually why you started shopping.

Question 4. Where is it processed, and by whom?

You need to be able to name the processor, the hosting region and the sub-processors. Not because a particular region is required, but because a privacy notice that cannot say where the data goes is a privacy notice you cannot defend. Ask for the data processing agreement before you sign up. A vendor who cannot produce one within a day is telling you something.

Question 5. Can you get your data out?

Ask on the way in, because the answer decides whether this is reversible. Is there a CSV export? A stats API? Raw event access? A tool with no export turns every future migration into a restart from zero.

Question 6. What shape is the pricing?

The number on the pricing page matters less than what it gets multiplied by. Three shapes are common and they behave very differently as you grow.

Traffic-based pricing matches what an analytics tool is worth to you: you pay more when there is more to measure. Per-seat pricing on an internal dashboard produces the same outcome everywhere, which is one shared login and a team that cannot see its own numbers.

Question 7. How long is the data kept?

Retention is where free tiers are quietly funded. Analytics that keeps two months of history cannot answer a year-over-year question, and you will want a year-over-year answer eventually. Ask for the number, and ask what happens to data past it.

Question 8. What happens to the history you already have?

It stays where it is. There is no meaningful path to move historical GA4 data into another tool, which leads to the one piece of migration advice that always applies. Install the new tool alongside the old one and run both for at least thirty days. You get a period where the numbers overlap, you learn how big the gap between the two definitions is, and GA4 keeps collecting until you are sure. The overlap is not wasted. It is the only calibration you will ever get.

The Google Analytics alternatives worth an afternoon

Short on purpose, fair on purpose, and it includes tools that beat us on things that matter.

• Plausible. Cookieless, no banner, no sampling, open source and self-hostable. Mature, with a large public customer base. For the pure web-analytics job it is the safest recommendation on this list and there is no argument otherwise.
• Fathom Analytics. The other strong cookieless option. API access on every plan, long retention and generous site allowances. If you run several small sites, that allowance alone can decide it.
• Matomo. The most feature-complete: heatmaps, session recording, goals, and the ability to import GA history. Self-hosted or cloud. Heavier than the others, and that is the trade.
• Umami. Open source and free if you host it yourself. You own the operations. If that appeals rather than appals, it is a real option.
• DataSpice. Ours. The difference is scope rather than web analytics: App Store Connect and Stripe revenue live in the same workspace as the traffic, so "which source produced paying customers" stops being a manual join. Against the list above we are new, we have no public customers, and we have no export or API yet. The row-by-row versions are in the comparisons.

Which of the eight questions to answer first

Question one settles most of the debate. Question three decides whether the banner comes off. Question five decides whether you can change your mind later. The rest is preference, and preference is allowed. You are going to look at this thing every morning, and a tool you enjoy opening gets opened.

If DataSpice is on your shortlist, the head-to-head is written out row by row in DataSpice vs Google Analytics 4, and it opens by saying what GA4 does better.