Mission 03

GA4 & GTM Implementation

GA4 and Google Tag Manager implementations that hold up under scrutiny. We design the event schema, build the data layer spec, instrument the site, and document it so the next person (human or AI) can make sense of it in under ten minutes.

Scope this out
Deliverables

What you get.

  • Event taxonomy and data layer specification
  • GTM web container with triggers, variables, and Consent Mode v2
  • GA4 configuration, custom dimensions, and enhanced measurement tuning
  • QA via Tag Assistant, GA4 DebugView, and end-to-end test plans
  • Documentation designed to be read, not ignored
Outcomes

Why it matters.

01

Analytics you can trust when the CFO asks how the number was built

02

Documentation that survives staff turnover

03

Privacy compliance that does not destroy measurement

Common questions

What people ask about ga4 & gtm implementation.

01 Do you follow GA4 best practices, or do you have a custom approach?

Both. We follow GA4's documented event schema where it makes sense, and deviate where it doesn't. For example, recommended events like purchase are used as-is. But enhanced measurement is often too noisy by default — we tune it. Our north star is always: can the next analyst trust this data without asking questions?

02 Can you fix an existing GA4 implementation, or do we need a fresh start?

Fix. We audit the current setup, document what's broken, and remediate in place. Starting fresh means losing historical comparability, which is rarely worth it. The one exception: if the current implementation has duplicate or inflated conversions flowing to Google Ads bidding, we pause the bad tags immediately and fix in parallel.

03 Do you set up Consent Mode v2?

Yes. Consent Mode v2 is mandatory for EEA traffic serving Google Ads, and we implement it as standard practice regardless of jurisdiction. We pair it with a compliant consent management platform (CookieYes, Cookiebot, OneTrust, or your choice), validate consent signals are firing correctly, and document fallback behavior for declined consent.

04 Will you document what you build?

Always. Every implementation ships with a tracking plan document: event schema, data layer specification, tag-to-event mapping, QA test cases, and a "how to extend this" section. The test is simple — if a new marketing hire couldn't open the doc and understand what fires where, the doc isn't done.

05 What about BigQuery export — do I need that?

If you're spending meaningfully on paid media ($25K+/month), yes. The GA4 UI has sampling thresholds and retention limits that will eventually bite you. BigQuery export is free (at GA4's reporting volume), retains raw event data forever, and unlocks analysis the UI can't do. We enable it as part of every implementation.

Let's see if we're a fit.

Tell us about the account. We'll tell you honestly what we'd do.

Start a project