If you manage a website, you probably had the same reflex as we did in our early days: installing Google Analytics. It's free, it's the market standard, and everyone uses it.
But while creating our latest side-project, we made an alarming observation looking at our first dashboards: our data was wrong. Worse, it was unreadable.
The phantom traffic mirage
By digging into our raw server data, we realized that a massive part of our traffic (sometimes up to 40%) did not come from humans, but from bots. Indexers, AI scrapers, automated scripts hosted in massive data centers (notably in the USA).
What does GA4 do about this? It does have a bot filter, but it's a 'black box'. It discards some bots from its official lists but lets thousands of more sophisticated automated scripts through, or those that might suit them, artificially inflating your visit statistics and mechanically destroying your conversion rate.
The decision-making factory
Beyond the accuracy of the numbers, we hit another wall: complexity. For my co-founder (non-tech profile), analyzing an 'engagement rate' crossed with a 'referrer' was a headache. Numbers remain numbers; they don't provide direction.
Our solution: Alternytics
Rather than paying for complex solutions or sacrificing our ethical values for free data, we coded our own solution.
- Transparent filtering: We isolate the bots and show them to you. No black box. You know exactly what is human and what is machine.
- AI for action: We integrated the European model Mistral AI. Every week, the AI analyzes your real human data and sends you a clear summary via email with 3 actions to implement.
"We stopped looking at our statistics. We listen to them."