Why estates make this move
- Container sprawl - per-brand GTM containers with divergent conventions and no shared taxonomy.
- Identity and audience requirements that a tag manager was never meant to carry.
- Consent enforcement that must be provable across client and server collection.
- A policy decision toward vendor-neutral infrastructure. (If none of these apply, keep GTM - see the comparison.)
The concept mapping
| GTM concept | Tealium equivalent | Migration note |
|---|---|---|
| Container | iQ profile | Multi-container estates usually consolidate to fewer profiles with shared standards |
| Tag | iQ tag | Marketplace templates cover the common vendors; custom HTML tags get re-reviewed, not copied |
| Trigger | Load rule | Rebuild from the event spec - triggers accumulated over years should not survive verbatim |
| Variable | Data layer attribute | The big win: variant dataLayer conventions converge on one specified object |
| dataLayer.push | utag data / event calls | Template-level change - inventory which pushes exist before promising a date |
| Server container (sGTM) | EventStream | Re-specify, do not port - the spec-driven model is the point of the move |
The order of work
Inventory the containers at runtime (what fires, not what exists - GTM containers are typically a third dead weight). Write the target event specification. Build iQ profiles to the spec while GTM keeps running. Reconcile destination by destination - GA4 first, usually, since its numbers are the ones leadership watches. Cut over per destination, then empty the GTM container deliberately. The migration hub covers the method in full.
Do not port the debt. The most expensive mistake in this direction is faithfully recreating years of GTM workarounds in iQ. The migration is the once-a-decade chance to rebuild the taxonomy clean - taking it is usually worth more than the platform change itself.