How Bizee Cleaned Up Years of GTM Bloat and Built a Marketing Data Foundation They Can Finally Trust

Bizee is a leading online business formation platform that has helped over a million entrepreneurs start and run their businesses. They know their product inside out, but analytics is a different kind of infrastructure, and it doesn't govern itself. By the time Bizee came back to Adasight for this engagement, their GA4 and GTM setup had been running for years without a dedicated owner. Tags had been added whenever someone needed something. Containers had grown without structure. Legacy code sat alongside new code, and nobody was quite sure what was still needed and what wasn't. The result was a marketing data layer that was technically live but practically untrustworthy, mand a team making spend decisions they couldn't fully stand behind.

This was Bizee's second engagement with Adasight. The first had fixed their Amplitude product analytics foundation — diagnosing zero-conversion funnels, resolving a broken identity model, and enabling the team to run experiments independently. You can read that story here. This time, the focus was the other half of the data stack.

‍

‍

• 16 of 17 Recommendations Shipped — Every High-Priority Item Complete

A full audit across 49 checkpoints produced 17 prioritized recommendations. 16 were executed within the engagement, with every high-priority fix shipped and only one low-priority item remaining.

• Broken Attribution Fixed

Referral traffic was landing in "(not set)" due to missing cross-domain tracking and referral exclusions. Both were configured, closing the attribution gap and restoring accurate channel reporting for Google Ads and paid campaigns.

• 221 Unfiled GTM Items Organized and Legacy UA Removed

Two GTM containers had accumulated hundreds of tags, triggers, and variables with no folder structure. All 221 unfiled items were organized, and legacy Universal Analytics tags were removed — eliminating double-tracking risk and making both containers maintainable.

• Clear Boundary Defined Between GA4 and Amplitude

With both tools in use, there was no documented answer to "where do we look for what?" Adasight produced a GA4 vs Amplitude boundary guide — GA4 and GTM for acquisition, ad-platform conversions, and web attribution; Amplitude as the source of truth for product analytics — giving the team a clear, shared framework for everyday reporting decisions.

• Two Teams Trained and Able to Maintain the Setup Independently

Two 45-minute live training sessions — GTM Maintenance and Tag Lifecycle, and GA4 Data Interpretation — were delivered with slides shared for ongoing reference, ensuring more than one person could confidently own the setup going forward.

‍

A marketing data stack that had outgrown its governance

Bizee had already fixed their product analytics. After an earlier engagement with Adasight that rebuilt their Amplitude foundation from scratch — diagnosing zero-conversion funnels, resolving a broken identity model, and enabling the team to run experiments — the product side of the data stack was in good shape. Marketing was a different story.

GA4 and Google Tag Manager were live and actively used for Google Ads conversion tracking and channel attribution. But the setup had grown organically over years without any governance to keep it clean. Two GTM containers had accumulated hundreds of tags, triggers, and variables — 221 items sitting unfiled in a single container alone. Legacy Universal Analytics tags were still running alongside GA4, creating a double-tracking risk nobody had formally addressed. Referral traffic was regularly landing in "(not set)" because cross-domain tracking and referral exclusions had never been configured.

The numbers existed, but nobody could fully trust them

The practical consequence was that the marketing team was making acquisition and spend decisions on data that was hard to validate. There was no separation between staging and production environments, no documented tagging specification, and no clear answer to where GA4 ended and Amplitude began. That last gap was particularly costly: with two analytics tools in use and no defined boundary between them, everyday reporting decisions — which tool to open, which number to trust — were harder than they needed to be.

Bizee needed a full audit, a prioritized fix plan, and the internal capability to keep the setup clean going forward.

‍

Starting with the full picture, before touching anything

The engagement started with a comprehensive audit rather than jumping straight to fixes. GA4 was reviewed against 49 checkpoints across 8 configuration areas, and both GTM containers were inventoried end to end. Every finding was categorized as working, broken, redundant, or missing — giving a complete, honest picture of the setup before a single change was made.

Turning findings into a plan everyone could act on

Each audit finding was turned into a single action item with a plain-English rationale, an impact and effort rating, a priority level, and a clear owner. The output was one shared plan of 17 recommendations — unambiguous and prioritized so the team knew exactly what to tackle first and why.

Drawing the line between GA4 and Amplitude

Before executing any fixes, Adasight documented the deliberate division of responsibility between the two tools: Amplitude as the source of truth for product analytics, and GA4 with GTM for acquisition, ad-platform conversions, and web attribution — with a deliberate overlap for cross-checking. This wasn't just a deliverable. It was the framework that would prevent the same confusion from creeping back in.

Executing the fixes that mattered most

The high-priority items were executed directly: cross-domain tracking and referral exclusions configured to close the "(not set)" attribution gap; legacy Universal Analytics tags removed; event names standardized; 221 unfiled GTM items organized into clear folders; a tagging specification created; and GTM staging environments set up to keep the setup reliable as the site evolves.

Leaving the team capable of maintaining it

Two 45-minute live training sessions were delivered — GTM Maintenance and Tag Lifecycle, and GA4 Data Interpretation — with slides shared async for ongoing reference. The goal was for more than one person to be able to confidently maintain and extend the setup without needing to bring in external support every time something changed.

‍

From bloated and broken to lean, trusted, and maintainable

Bizee entered the engagement with a marketing data setup that had outgrown its governance — bloated containers, broken attribution, legacy tracking still running, and no shared framework for how GA4 and Amplitude should work together. By close, all of that had been resolved.

The fixes that moved the needle

16 of 17 recommendations were shipped within the engagement, with every high-priority item complete. Referral attribution was restored, double-tracking risk eliminated, and both GTM containers organized into a structure the team can actually navigate and maintain. The GA4 vs Amplitude boundary guide gave everyone a clear answer to "where do we look for what?" — removing the friction that had been slowing down everyday reporting decisions.

A foundation that connects both sides of the data stack

Two teams were trained across GTM maintenance and GA4 interpretation, with documentation in place to support them going forward. Bizee now has a leaner, more accurate marketing data foundation — one that connects cleanly to the Amplitude product analytics foundation built in the earlier engagement, and that the team can trust and manage independently as the business grows.

Want the same confidence in your marketing data?

• Data Stack Audit — diagnose what's broken, missing, or duplicated in your existing GA4 or GTM setup before you build on it
• Data Mapping & Integration — connect your data sources and get clean, reliable data flowing into your analytics stack
• Data Foundation — tracking plan, taxonomy, dashboards, and governance built around your actual use cases

Book a call with Adasight and we'll help you get it right.

‍