Hi, I'm Paolo. MarTech Engineer and Tracking Specialist. I work with data in marketing since 2018 and I like technology, coding, privacy and learn how the world works.
I lift heavy circles, cook and make bread.
The blog is in English, sometimes in Italian.
Three domains, six advertising platforms, and years of accumulated tracking debt. How I redesigned the data layer, consolidated everything into GTM, and restored reliable conversion data across GA4, Meta, Google Ads, TikTok, LinkedIn, and Microsoft Ads.
Server-side vs client-side GTM compared: +39pts PageSpeed, 24-36% more conversions, $100+/mo cost. Real data and decision framework for 2026.
A real-world audit of Adobe Analytics data layers across 8 European markets. Comparing digitalData vs adobeDataLayer, page type taxonomies, product data population, and key takeaways for multi-market implementations.
Open any GA4 report and look at Total Users. Then run COUNT(DISTINCT user_pseudo_id) in your BigQuery export. The numbers won't match. They're usually within a percent or two, and there is no setting that makes them. That gap isn't a bug, and it isn't sampling. It's a deliberate engineering decision: GA4 never actually counts your users. It estimates them, using a probabilistic data structure called HyperLogLog++ that can approximate the number of unique items in a set of any size using a fixed 12 KB of memory.
I design and fix measurement setups: data layers, GTM web and server-side containers, GA4 and consent. I care about how the data flows underneath, not which button to click.
Lately I'm also learning C, which is where the hexdump and Game of Life posts come from.
Need your tracking audited, rebuilt or explained?