Google Ad Manager is the platform you use when the ads on your site are not yours: an ad server that arbitrates between advertisers competing for your audience. Full guide: how it differs from Google Ads and the MCC, the building blocks, the ad request lifecycle, every configuration point, and a field-tested debugging toolkit.
Read Article →Three Google products share the Looker name. Two are products. One is a language. All three do fundamentally different things, serve different audiences, and cost very different amounts. This article maps them clearly — including when they work together.
Read Article →The most complete guide to Looker Studio (formerly Google Data Studio): interface walkthrough, all connector types, every chart, calculated fields with real formulas, data blending step by step, parameters, pro tips, honest limitations, and a detailed comparison with Power BI and Tableau.
Read Article →30% of your SFCC orders have no GA4 session. This is not a bug. It is a mix of five structurally different categories — each requiring a different response. This guide covers every category in depth, the compounding analytics errors they cause if ignored, a two-tier LookML classification system, all the metrics to build on top, and the complete reconciliation dashboard architecture.
Read Article →Five dashboards, five revenue definitions, one team. This is not a data quality problem — it is a governance problem. This guide covers the full LookML governance approach: the right way to define dimensions and measures, cross-source metrics, derived tables, Git versioning, access control, and the real-world workflow that prevents metric drift permanently.
Read Article →Your LEFT JOIN between GA4 and SFCC returns 40% NULLs. The problem is almost always the transaction ID. This article covers every failure mode with real examples, a decision-tree for diagnosing your specific case, four SQL queries to run before touching Looker, three LookML fix patterns, and a benchmark table for what healthy match rates look like by segment.
Read Article →GA4, SFCC and your offline files will always show different revenue numbers. This is not a bug. Each system measures a different reality, at a different moment, with a different definition of what counts. This deep-dive covers every structural cause, the full data flow architecture, practical BigQuery diagnostics, and the business questions that only become answerable once you understand why the numbers differ.
Read Article →Looker is not a dashboard tool. It is a semantic layer platform built around LookML — a modeling language that defines your business logic once and enforces it across every report, every user, and every team. This guide covers everything: architecture, LookML fundamentals, joins, calculated metrics, derived tables, access control, scheduling, performance, and real-world use cases.
Read Article →Introduction Server-side tracking is often presented as a universal fix for: The common belief is simple: “If tracking runs on the server, browsers can’t interfere.” This is incomplete. Browsers do not block based on where your container runs.They apply restrictions based on: Server-side tracking is not a shield.It is an architectural shift. And architecture can […]
Read Article →Cookieless is not determinist Clearing cookies used to feel like a reset. It is not. Modern tracking increasingly relies on measurement (fingerprinting) and side-channels (caches and protocols people rarely think about). To make this concrete, I tested my own browser with the EFF’s Cover Your Tracks tool. The result was a unique fingerprint among 303,650 […]
Read Article →Introduction The term “cookieless” has become the hottest buzzword in digital marketing and ad tech. You can’t attend a conference, read a vendor pitch, or scroll through LinkedIn without seeing someone touting their “cookieless solution.” Google keeps postponing the deprecation of third-party cookies in Chrome. Safari and Firefox have already blocked them. The GDPR and […]
Read Article →Three containers. Perfect architecture on paper. 60% unset variables in production. That was my week. A client approached me with what seemed like a textbook Tag Management System setup: one core container for essential tracking, two business-unit-specific containers for specialized marketing tags. Clean separation of concerns. Different teams managing their own implementations. Excellent governance model. […]
Read Article →