Uncategorized
Google Analytics alternatives worth a look
What the main alternatives do differently, and what you lose by moving. Then the one reason to switch that has nothing to do with disliking the interface.
Correction, 23 September 2026: the earlier version overstated server-log completeness and treated cookieless tools as generally exempt from consent. Both claims are now qualified.
Choose analytics by the question you need to answer. Browser collection can miss visits when scripts are blocked or consent is declined.
If that is your problem, the interface you move to matters less than whether the measurement happens somewhere a blocker cannot reach.
What you actually lose by moving
Ad platform integration. If you run Google Ads, the analytics link supplies conversion data and audiences back to the auction, and replacing that is real work.
Free scale. Google Analytics is free at volumes where paid tools cost real money.
Historical data. Nothing migrates cleanly. You start a new series, and for a year you are comparing across two systems. Run both in parallel during that period, which everyone regrets skipping.
The categories
Lightweight web analytics. Plausible documents its cookieless approach. Cloudflare Web Analytics documents its own collection and reporting. Compare the actual fields and deployment settings. A product category does not determine your consent obligations.
What you give up is depth. There is no funnel exploration, no cohort analysis and no session replay. For a marketing site, that is often the right trade, because nobody was running those reports anyway.
Product analytics. Event analysis helps answer questions about activation and repeat use. PostHog’s repository describes its self-hosted option and operating limits. Check those limits before choosing it for a production workload.
These are heavier, both in script weight and in the work of defining an event taxonomy. Done badly they produce a thousand events nobody can interpret.
Log-based and server-side. These records cover requests that reach the server or edge where logging is configured. They can include bots and miss cached requests at another layer. They do not describe every action inside a page. Collection still needs an appropriate privacy and consent setup.
Compare server records with browser events for the question at hand. Document the exclusions in each before choosing a reporting baseline.
Matomo offers an on-premise edition. Owning the deployment also means owning its updates and backups. Compare the reports you need before moving a working setup.
How to choose in ten minutes
If you need Google Ads conversion data, keep Google Analytics and add something else alongside it.
If you have a marketing site and want to know which pages and sources produce enquiries, take a lightweight privacy-first tool and stop there.
If you are measuring a product with sign-ups and retention, take a product analytics tool and spend the time on the event taxonomy.
If your audience blocks scripts, count server side and treat everything else as a sample.
The part people skip
Whatever you choose, define the one action that matters before you install it. A tool cannot tell you whether the traffic was any good if nobody has said what good means.
We wire measurement as part of conversion work, and the first question is always that one.
The ICO’s storage and access guidance explains why implementation and purpose matter when assessing tracking technologies.