GTM analytics · Multi-product Shopify SaaS
I turned organic, inconsistent website tracking into a reusable framework that could distinguish navigation from buyer intent, attribute every action to its source and scale across future marketing assets.
01 / The before
Tracking had scaled the same way the marketing website had: organically, one button at a time, without a shared architecture underneath it.
“Which pages and which CTAs actually drive demo bookings?” was a business question the analytics property could not answer.
02 / Measurement architecture
Website audit → customer journey → CTA inventory → event taxonomy → parameter design → implementation → QA → validation
Where visitors act—and where each action becomes measurable.
Homepage, blog, product or industry page
cta_click · signalReads content and compares products
cta_click · signal47 actions audited and classified
intent-based routingEvent fires with three dimensions
key event or signalDemo booked or lead form submitted
GA4 key eventNew buttons inherit a category; they do not create a new reporting language.
Demo or discovery-call clicks
Lead form successfully submitted
Blog sidebar and footer CTAs
Navigation and secondary CTAs
Every event carries the same three dimensions
Location-level attribution without a new tag for every button.
hero | footer | bottom_bannerWhich position drove the action?Book a demo | Book a discovery callWhich wording performed?global | page_specificWas it site-wide or contextual?Nothing shipped until both GTM Preview and GA4 Realtime agreed.
I documented what was tested and flagged the row instead of guessing. A tag stayed unpublished until Preview and Realtime both confirmed it.
03 / Decision anatomy
Website tracking had grown one button at a time across four product lines. Fragmented names and missing conversions meant the business could not reliably identify which pages and CTAs drove demo intent.
The problem was not simply incomplete tagging. There was no shared definition of conversion intent, no reusable metadata model and no QA standard for future marketing assets.
Replace one-off button tracking with an intent-based event taxonomy, a shared three-parameter attribution model and a repeatable validation workflow.
Customer-journey mapping, a 47-CTA audit, conversion classification, event and parameter design, GTM implementation, QA, GA4 validation and documentation.
Marketing leadership and web, analytics and publishing stakeholders—to validate CTA intent, confirm destinations and close the gap between measurement design and live implementation.
Four reusable event families, three standard attribution parameters, 11 GTM tags and triggers, a verification register and a documented six-step QA procedure.
Restored measurement of the primary demo-booking action, standardized CTA reporting and gave future pages a framework they could inherit without rebuilding the tracking logic.
The company moved from counting isolated button clicks to classifying buyer intent consistently. Measurement became a system the website could scale, rather than a collection of tags only one person understood.
04 / What changed
This project did not claim downstream revenue impact it could not yet measure. Its result was the reliable measurement layer required to make future conversion and content decisions defensible.
Restored measurement of the primary demo-booking action.
Replaced button-level fragmentation with one intent-based taxonomy.
Made page location, wording and CTA type available for analysis.
Created a QA standard another team member could repeat.
Surfaced missing signup and trial conversions as an infrastructure gap requiring follow-up.