documentation

How Google Tag Manager behaves

Reference pages on how GTM actually works — written from measurements of real containers, not from the documentation. Each one states plainly what a published container can tell you, and what it cannot.

Server-side GTM (sGTM): what it is, how it works, what it takes to runA second GTM container, running on your infrastructure, that receives measurement data from the browser and decides what to forward to each vendor.
Custom HTML tags: the code in your container nobody reviewedA Custom HTML tag is arbitrary JavaScript, delivered to every visitor, deployed by whoever has container access — with none of the review your application code goes through.
Third-party hosts: who your container lets into the pageA container's third-party hosts are the list of organisations you have delegated execution to. Each one can change what it serves without telling you.
Consent Mode: how it works in the container, and why absence proves nothingConsent Mode is visible in a container in four different forms — and the most common way to implement consent is not visible there at all.
Reading the container mapEvery report flattens a container into lists and throws away the thing that makes it hard to reason about: what depends on what.

Why these exist

Most of what is written about GTM describes the product as intended. These pages describe it as observed, across 42,089 published containers — including the parts that are inconvenient, like the several places a setting can hide, and the questions no artifact can answer no matter how carefully you read it.

When we get one wrong we say so on the page rather than quietly editing it. A reference that cannot be checked is just an opinion with better typography.

Start with a real container. No login, no signup.

Run a container check