When a national home security company came to Pam Ann Marketing, they had a problem that is more common than most marketing teams realize: their Google Analytics 4 data was mixing two completely different audiences into one view. Every funnel metric, every conversion rate, every channel attribution number in their reports was partially wrong, and no one on the team knew it.
This GA4 case study walks through what we found, what we built, and what the marketing team could finally see once the data foundation was solid.
- Industry: Home Security Services
- Client Type: National B2C Brand
- Engagement: Ongoing Analytics Retainer
- Timeline: 18 Months
- Core Services: GA4 Audit, GTM Implementation, Looker Studio Reporting
The Problem: Data That Doesn’t Mean What You Think It Does
This client’s website served two fundamentally different audiences. Prospective new customers visited to learn about home security plans and request quotes. Existing customers visited to manage their accounts and reach support.
Both audiences were landing in the same Google Analytics 4 property. That meant every metric the marketing team used to evaluate performance (engagement rate, sessions, goal completions, channel attribution) was blended across people who were shopping for the first time and people who already had an account.
The real-world impact: If a prospect visits, checks service availability in their zip code, and leaves because coverage isn’t available in their area, that looks like a bounce or a low-quality visit in the data. If an existing customer logs in to pay a bill, that looks like an engaged session, even though it has nothing to do with marketing performance. Mixing these together makes every number less meaningful.
On top of the segmentation issue, the team lacked the granular conversion tracking needed to understand their actual funnel. They couldn’t see which specific forms were generating leads, whether phone calls were being counted as conversions, how many people were abandoning the service-area check process, or where form friction was causing drop-off.
Why Mixed Analytics Data Is So Common
This situation is not unusual. Many brands that serve both prospective and existing customers end up with GA4 setups that were originally designed for one audience and were never updated as the site evolved. When a company adds a customer portal, a support section, or call-center tools to the same domain as its marketing website, the analytics setup rarely gets updated to match.
The result is a data problem that’s invisible unless you know what to look for. Engagement rates look artificially high. Conversion rates look inconsistent. Channel attribution is off because returning customers are inflating the direct traffic numbers.
In this client’s case, a third-party technology audit had also flagged their analytics setup as problematic, but without understanding the intent behind the configuration. Part of our work was clarifying what was deliberate, what genuinely needed fixing, and what the tradeoffs were for each option.
What We Did: Our GA4 Audit and Implementation Approach
Step 1: Full GA4 and GTM Audit
We started by documenting every tag firing in Google Tag Manager and every event appearing in GA4. The goal was a complete map of what was being tracked, what was misfiring, and what was missing entirely. This audit became the foundation for all implementation decisions going forward.
During the audit, we identified the primary data integrity issue (prospect and existing-customer traffic in the same property), catalogued incomplete conversion tracking, and found several events that were firing incorrectly or duplicating data.
Step 2: Data Segmentation Strategy
We designed a three-property GA4 architecture to cleanly separate the data by audience:
- A www-only property for marketing site traffic, which is the view the marketing team uses to evaluate acquisition performance
- An existing-customer property for account-holder sessions
- A combined property for stakeholders who need the full picture in one view
The tradeoff with this approach is that new tracking configurations need to be deployed to multiple properties rather than one. We documented that tradeoff, weighed it against the clarity it provided, and set up a process to keep the properties in sync as new events were added.
Step 3: Custom Conversion Tracking Build-Out
With the segmentation architecture in place, we built out a comprehensive event taxonomy using Google Tag Manager. The goal was to track every meaningful action in the prospect funnel, not just form submissions, but the micro-behaviors that precede or follow a conversion.
Key tracking we implemented included:
- Form submissions by type: residential leads, commercial leads, and form completions segmented so each category could be analyzed independently
- Service area checks: a custom event to track when visitors check whether coverage is available in their zip code, including a separate event for when coverage isn’t available (a previously invisible drop-off point)
- Form error tracking: an event that captures the specific validation message shown when a form submission fails, giving the team diagnostic data on where and why people abandon forms
- Click-to-call tracking: categorized by placement on the page so call volume by location could be analyzed
- CTA click tracking: including HubSpot marketing banners and pop-ups that are embedded in iframes and are not tracked by GA4 by default
- Navigation interaction tracking: clicks on key menu elements and quote CTAs
Step 4: Call Tracking Integration
The client received a significant volume of phone inquiries alongside form submissions. Without call tracking connected to GA4, those conversions were invisible in the data. We integrated call tracking metrics (including first-time caller events and call duration thresholds) so phone and form conversions could be analyzed together in the same reports.
Step 5: Looker Studio Reporting Dashboard
We built a Looker Studio dashboard connected to the GA4 data so the marketing team and leadership could view funnel performance in a consistent, shareable format. The dashboard was designed around the specific questions the team was trying to answer, not a default report template. Because Looker Studio reports are shared directly with stakeholders, team members could review performance without needing individual access to GA4.
Step 6: Ongoing Monitoring and Advisory
Throughout the 18-month engagement, we met with the client’s marketing team every two weeks to review findings, investigate data anomalies, and prioritize the next phase of work. That regular cadence meant issues were caught and addressed quickly, and the roadmap stayed aligned with what questions the team was actually trying to answer.
Services Delivered
- GA4 Audit: Complete review of existing tracking setup, identifying what was accurate, misfiring, or missing.
- Data Segmentation Strategy: Multi-property GA4 architecture designed to separate prospect and existing-customer traffic cleanly.
- GTM Implementation: Custom event taxonomy built in Google Tag Manager, covering all meaningful prospect funnel actions.
- Call Tracking Integration: Phone conversion data connected to GA4 so calls and form leads could be analyzed together.
- Looker Studio Dashboard: Custom reporting dashboard built for the marketing team and leadership, shared directly with each stakeholder.
- Ongoing Analytics Advisory: Bi-weekly working sessions covering data review, anomaly investigation, and strategic prioritization.
Results: What the Team Could Finally See
Over 18 months, the engagement moved this client from unreliable blended data to a clean measurement foundation built around the questions that actually drive marketing decisions. Here is what changed.
- Trustworthy funnel metrics: Marketing performance data now reflects actual prospect behavior, not a blend of new visitors and returning customers.
- Service-area drop-off quantified: For the first time, the team could see how many visitors checked coverage availability and left because service wasn’t in their area (a significant, previously invisible funnel gap).
- Residential and commercial leads split: Two distinct lead types tracked separately, enabling smarter analysis of which channels and campaigns drove each.
- Phone conversions counted: Call data integrated into the same reporting view as form submissions, giving a complete picture of how inquiries came in.
- Form friction identified: Specific validation errors and form abandonment points surfaced as data the team could act on.
- Reporting the team actually used: A shared Looker Studio dashboard built around the team’s real questions, reviewed together on a regular cadence.
Key Takeaways for Your Organization
This engagement highlights a few patterns we see consistently across brands with complex websites and mixed audiences.
Data quality is a prerequisite for optimization
Conversion rate optimization, ad spend decisions, and channel strategy all depend on the data being accurate first. A GA4 audit is not just housekeeping. It’s the step that makes every downstream decision more reliable.
What looks like a tracking problem is often a segmentation problem
Many analytics issues we investigate turn out to be caused not by broken tags but by data that was never separated in the first place. If your GA4 property includes audiences who have nothing to do with your marketing funnel, your funnel metrics are not telling the truth.
Granular events reveal what aggregate metrics hide
Aggregate conversion rates can look stable even when specific steps in the funnel are broken or underperforming. Custom event tracking at the micro-behavior level (form errors, service-area checks, CTA clicks by placement) is what reveals where the real friction is.
Regular review keeps the data useful
Analytics setups drift over time. Websites change, forms update, new tools get added. The bi-weekly advisory cadence in this engagement meant issues were caught and corrected quickly, rather than going unnoticed for months.
Ready to find out what your GA4 data is actually telling you? Pam Ann Marketing specializes in analytics audits and implementation for brands that need measurement they can trust. Get in touch to start a conversation.
- How to Optimize Your Article Content for SEO and GEO/AEO - August 28, 2026
- Why All AI-SEO Studies are Flawed (and What to Trust Instead) - March 12, 2026
- How Much AI-Generated Content is Acceptable for SEO Writing? - February 25, 2026