Notes from the privacy-first stack.
Practical writing on privacy-first web analytics, what GA4 actually measures, and how AI changes the reporting workflow. From the engineers and operators building datataste.
How to track users across multiple domains and subdomains without losing attribution
A practical implementation and troubleshooting guide to cross-domain and subdomain tracking for organizations whose web estate spans separate domains.
Measuring marketing when your sales cycle is six months a practical framework for B2B teams
A repeatable way to decide what to measure before configuring a single tag — built for long, multi-touch, partly-offline sales cycles.
Stop waiting on the analytics team how self-serve analytics gives marketing its own answers
A day-in-the-life look at how non-technical marketing and campaign teams get answers from their website data — without writing queries or filing a ticket to a central data team.
The agentic web is here: how AI agents and bot traffic are quietly corrupting your analytics
AI agents and automated browsers are now visiting websites to act on behalf of users, and the broader bot era is reshaping web traffic. This is what it does to your numbers — and how to keep them trustworthy.
How website changes silently break your tracking — and how to catch them before your data lies
Most broken tracking isn’t a consent problem — it’s a side effect of a routine site change. Here’s how those failures happen, why they stay invisible, and how to build a process that catches them.
Why GA4 shows you far less than 100% of users — and how privacy-first tools close the gap.
Consent banners, data modeling, browser restrictions, ITP, ETP, and the legal constraints behind every "(other)" row in your GA4 reports. We walk through where 30–60% of visitor data disappears and what a privacy-first architecture actually does differently.
How to design a tracking setup that works without consent.
Step-by-step: architect analytics that collects no personal data — no cross-site IDs, no raw-IP storage. Includes a field-by-field audit table and DPIA template.
Why server-side tracking alone does NOT fix your data quality.
Server-side GTM still needs cookies, consent, and client-side JS. We break down why it's not the silver bullet agencies sell it as.
How Google Consent Mode actually works (and why it still loses data).
Technical teardown of GA4 Consent Mode v2: cookieless pings, data modeling, selection bias, and the ad-blocker blind spot.
One email. The new issue plus a single chart we couldn't fit.
No marketing. Unsubscribe in one click.