Know where they come from
Search, AI assistants, campaigns and the sites that link to you, separated out, with the visitors each one actually brought. You can see whether ChatGPT is recommending you rather than guess at it.
Built in-house
Completely cookieless and GDPR compliant by design. A web-analytics platform I built from scratch. It answers everything Google Analytics answers, while storing no personal data, setting no cookies, and requiring no consent banner on your site.
Cookie banners exist because scripts store things on a visitor’s device. This one stores nothing: no cookies, no local storage, no session storage, so the rule that mandates the banner is never triggered.
Analytics normally means handing your visitors to a third party. Here nothing is handed anywhere: the whole platform runs on infrastructure I control, and not a single request goes out to an advertising network, a data broker, or a third-party CDN.
Every event is kept in full and can be queried directly. No sampling thresholds, no aggregation you cannot see behind, and no retention limit imposed by someone else’s pricing tier.
Which companies visited. Which page loses people. Which query brings clicks. Which backlink broke. Those are product decisions, not settings in someone else’s tool.
What you can see
vs. previous 7 days
Search, AI assistants, campaigns and the sites that link to you, separated out, with the visitors each one actually brought. You can see whether ChatGPT is recommending you rather than guess at it.
visitors
Every session reconstructed in order: the pages they opened, how long they stayed on each, and the point where they stopped. Nothing is sampled away, so a session worth looking at is one you can actually open.
Outbound clicks, downloads and form submissions, each with its exact destination. When someone leaves for your booking link, that is recorded as an outcome rather than written off as a bounce.
Destination
https://cal.com/fabrice/intro-call
Cookieless analytics is easy to claim and harder to do. Here is the entire identity mechanism, start to finish.
A visitor loads a page. Their IP address arrives with the request, as it must for any web request to work at all.
The IP is read to derive an approximate location, the organisation behind the network, and a hash salted with a key that changes daily.
The raw IP is discarded and never written to disk. Because the salt rotates every day, yesterday’s hash cannot be matched to today’s visitor.
The result: sessions can be counted accurately within a day, and nobody, including me, can reconstruct who a visitor was afterwards.
Not compliant because of a setting you have to remember to switch on, compliant because the data that creates the obligation is never collected in the first place.
Consent banners are mandated by the ePrivacy rule on storing or reading information on a visitor’s device. This platform writes nothing: no cookies, no local storage, no session storage, so the obligation never arises.
IP addresses are processed in memory and discarded, never written to disk. With no identifiers stored, there is nothing to export, rectify or erase on request, and nothing to lose in a breach.
Your analytics data is never shared with an advertising network or a third-party provider. It stays on infrastructure you control, which removes the international-transfer question entirely.
Everything lives in a database you own, with no retention limit dictated by someone else’s pricing tier and no vendor standing between you and your own numbers.
Worth saying plainly: software can remove the hard part, but compliance also depends on your own privacy notice and how you use the data. This platform is built so the technical side is not the thing standing in your way.
What this means for you
Most analytics ends with a dashboard nobody opens. The platform is the instrument. The useful part is somebody actually reading it. That is the part I do.
Nothing sampled away, so the month is complete before I look at it.
Plain language, not a dashboard export with a logo on it.
Ranked by what they are worth against what they cost to fix.
What moved
Worth doing next
Every month, against your actual numbers, not a generic checklist.
High exits, short visits, funnel steps where visitors quietly disappear, the specific page and step, not a site-wide bounce rate.
Core Web Vitals measured on real visits, per page. A slow template hurts both conversion and search position, and it rarely shows up on your own fast connection.
JavaScript errors and 404s captured from actual sessions. A form failing on one browser is invisible until someone reports it, which mostly they do not.
Queries with impressions but few clicks, or sitting just off page one. Usually the cheapest wins available: a title and description rewrite on a page that already ranks.
Crawled-but-not-indexed URLs and sitemap errors, so pages you paid to produce are not sitting invisible.
Dead pages on your own site that still attract backlinks. A redirect recovers authority you already earned and are currently throwing away.
UTM sources bringing traffic that never converts, attributed across the whole session rather than just the landing page.
Which organisations visited, how often, and what they read: a warm list for B2B, from a signal most analytics never surfaces.
The dashboard is still there if you want it. Nothing depends on you opening it.
BTOP Analytics runs on this very website. The numbers behind this page come from it. If you want measurement you own outright, tell me about your setup.