Google Tag Manager is a free tool from Google that runs in the visitor's browser, letting a team add and manage marketing and analytics tags without editing a website's code. It works through one container per site, holding the tags, the triggers that fire them and the variables that feed them data, published to the live site as a single snippet. That snippet only adds the container, and nothing is measured until tags are configured and published inside it.
Why it matters for agencies
For an agency, Google Tag Manager is usually the layer already running before any specialist gets involved, since most sites ship with the container installed by a developer or a previous agency. What sits inside that container, not the fact that it exists, decides whether a dashboard can be trusted. A client can point at an existing container as proof tracking works, when it might hold no published tags at all, or tags configured once and never revisited as the site or the ad accounts changed.
What teams get wrong
The most common mistake is treating the container as the tracking itself. Google's own documentation describes the container as holding the tags, triggers and variables used to measure interactions, and dispatching that data as HTTP requests once those pieces are set up and published, work the installed snippet does not do on its own. A site can run Google Tag Manager for years with an empty or half-built container and look, from the outside, exactly like a site with working measurement.
The second is confusing this container with the server-side product carrying the same name. Server-side Google Tag Manager runs its own separate container on infrastructure a team hosts, receiving what this browser container sends rather than running in the visitor's browser at all, and the two do different jobs. Archon Labs treats an inherited container as a starting point to audit, not a checkbox already ticked, before trusting anything downstream of it.
In depth: Archon Signal