The problem

GA4 was installed, not designed. Someone pasted the tag, checked the box, and moved on — so events fire twice or not at all, “conversions” count form views instead of submits, and the numbers shift every time the site changes. The result is a dashboard nobody trusts, which quietly becomes one nobody uses. And spend keeps flowing against data no one will stand behind.

What analytics & tracking is

GA4 and GTM rebuilt on purpose instead of by default: a small, deliberate event model — a handful of canonical events that mean the same thing everywhere — fired from stable data attributes baked into the markup, with conversion definitions mapped to what the business actually cares about. The difference between analytics that happen to you and analytics you designed. Built this way, a redesign or a vendor swap doesn’t silently break your measurement.

Who it is for

• You migrated to GA4 and lost confidence in the numbers
• Your “conversions” don’t match what actually happens in the business
• Events fire inconsistently — double-counted, missing, or on the wrong thing
• You’re about to increase spend and refuse to do it against data you don’t trust

How it works

Tracking done right is a plan first and a tag second:

• Define the canonical events — a short list that means one thing across the whole site (a lead is a lead, everywhere)
• Write the tracking plan — every event, its parameters, and where it fires, documented before anything is built
• Stable data attributes in the markup — hooks that live in the site contract, not in CSS classes that change every redesign
• Rebuild GTM from that plan — triggers and tags that match the document, not a pile of ad-hoc tags no one remembers
• Define conversions and audiences — mark the events that matter as key events so reporting and ad platforms optimize toward the right actions
• Verify before trusting — test each event in preview and debug tools until the data is provably correct

Measurement you can defend, not a dashboard you hope is right.

Included

• GA4 property and event architecture
• Written tracking plan (events, parameters, per-page)
• GTM container rebuilt from that plan
• Stable data-attribute tracking in site markup
• Conversion and audience definitions
• Verification in preview and debug before handoff

Not included

• Vanity dashboards without decisions attached
• Leaving tracking hostage to CSS classes and ad-hoc tags

What you walk away with

Analytics you can act on, that survive redesigns and vendor changes.

• A documented event model anyone can read
• Conversions that match revenue, not page views
• Tracking vendors cannot silently break
• Data you’re willing to make budget decisions on

Frequently asked questions

Why use data attributes instead of click triggers on classes?

CSS classes change with every redesign and break tracking silently. Stable data attributes are part of the site contract, so measurement survives change.

Can’t I just use GA4 Enhanced Measurement?

Enhanced Measurement is a fine floor for pageviews and scrolls, but it can’t see the actions that matter — a booking inside an embedded widget, a form that posts to your CRM, a phone click you don’t want double-counted. Those need deliberate events.

Will this survive a website redesign?

That’s the whole point of the data-attribute approach. Because the tracking hooks live in the markup contract rather than the styling, a redesign changes how the site looks without breaking what it measures.

Do you rip out my existing GA4?

Usually not. We keep the property and its history where it’s sound and rebuild the event model and GTM container around it, so you don’t lose continuity while gaining trust.

Related

• Revenue Attribution — Where clean events become revenue truth.
• Growth Systems Diagnostic — The starting point for every engagement.
• Growth Systems — The measurement, acquisition, and conversion layer behind growth.
• Fractional CMO — Marketing leadership, budget ownership, and operating cadence.