Google Tag Manager
What it is
A container that holds your tracking tags, so marketing ships or changes measurement without a code release.
What it’s good at
Speed and reversibility. A new tag is live in minutes; every change is versioned, so when a number turns strange you can read what changed and roll it back. It keeps tracking work off the developers’ release calendar.
What it costs to own
Minutes to install; a standing discipline to run. Every tag is third-party code on your pages: page weight and a privacy decision someone must be able to defend. Containers accumulate; agencies add tags, and cancelled tools leave theirs behind, still firing. The bill is an owner with sole publish rights and a yearly pruning review, where each tag answers what it collects and who reads the result.
When it’s a poor fit
It fits poorly when access is broad and ownership absent: if most of the team can publish, version history will only tell you who caused the incident. It also fits poorly as a way for marketing to route around a slow development team, since injected banners and patched pages are site changes made outside review. And when the whole setup is four tags and releases ship weekly, hard-coding them is often the simpler path, one less system to own.
What it sits beside
It deploys GA4 and the ad pixels, with your consent platform ahead of it so tags fire on the visitor’s choice. The classic overlap: a shop plugin fires the same pixel the container fires, and every conversion counts twice. The usual fix is one route per pixel, written down.
My experience
In most Tag Manager containers I have inherited, the first job has been deleting tags nobody used any more.