Repairing dynamic number insertion, GA4, and Google Ads integrations to create cleaner conversion data for reporting and optimization.
The most dangerous tracking setup isn’t one that produces no data. It’s one that produces plausible, but incorrect, data.
A business running marketing across several websites, locations, and phone numbers can carry this problem for years without noticing, because every report keeps looking fine. That was the situation we inherited at the beginning of a new engagement with a behavioral healthcare organization operating 17 separate treatment facility websites.
CallTrackingMetrics sat at the center of it: capturing calls, receiving form submissions, passing conversions into GA4 and Google Ads, and syncing lead data with Salesforce, all across a network of more than 1,000 tracking numbers. On paper, the structure made sense. In practice, years of separate implementations and site changes had made the system nearly impossible to verify.
Some calls and forms were counted twice. Others never reached GA4 or Google Ads. Conversion paths differed from site to site. Most importantly, the mechanism responsible for attributing phone calls wasn’t working as intended.
At a Glance
- 17 websites
- 1,000+ tracking numbers across online and offline sources
- 17 separate GA4, GTM, and Google Ads setups
- Calls and form submissions used as primary Google Ads conversions
- Dynamic number insertion and website phone-number cleanup across the network
What We Found
In the CTM account documented during the audit, more than 1,000 tracking numbers existed. Roughly 500 were configured for onsite use, meaning they were intended to participate in dynamic number insertion.
Only 60 had target numbers assigned to them. As a result, the remaining ~440 numbers intended for online use weren’t configured to dynamically swap. The tracking numbers weren’t being used, and the client was missing out on valuable activity and attribution data.
We also found:
- Source-specific tracking numbers hard-coded onto websites
- Numbers associated with the wrong facility
- Missing fail-safe numbers
- Inconsistent tracking sources and account mappings
- Duplicate and missing conversion paths into GA4 and Google Ads
Calls were still coming through. Reports were still populating. Sources were still appearing next to leads.
The problem was that the attribution behind those reports couldn’t be trusted.
The Problem Started on the Websites
Dynamic number insertion, or DNI, allows a website to display a different tracking number based on how someone arrived.
A visitor from a Google Ad might see one number. Someone coming through organic search, social media, referral traffic, or another source might see another.
When the person calls, CallTrackingMetrics can connect that inquiry to the source and website activity that preceded it.
For that process to work, CTM needs to know which permanent phone number on the website it should replace. That permanent number is called the target number.
Without a defined target number, CTM can’t determine what to swap. The call may still connect, but the platform can’t display the correct source-specific tracking number or attribute the inquiry accurately.
The organization wasn’t losing every call. It was losing clarity on where those calls came from.
The website audit uncovered another source of misattribution: tracking numbers had been entered directly into site content.
Some were assigned to a particular marketing source. Others were associated with a different facility.
Imagine that an organic search tracking number is hard-coded into a facility website. A visitor clicks through from Facebook, sees that number, and calls.
The call connects. CTM records it. A source appears in the report.
But because the number itself is assigned to organic search, the call is attributed to organic search instead of social media.
The data looks legitimate, but it’s wrong.
Auditing the Entire Conversion Path
Checking each piece in isolation wouldn’t tell us much; the whole path had to be tested as one system. To do this, we traced each call from the number on the site through attribution, facility routing, analytics, advertising, and CRM, and helped map the CTM–Salesforce integration so lead data stayed synchronized across both.
Rebuilding the Tracking System
Once we understood the environment, we began repairing and standardizing it.
The work included:
- Establishing facility-specific target numbers
- Adding and validating fail-safe numbers
- Removing hard-coded tracking numbers from website content
- Replacing them with facility-specific target numbers that DNI could reliably swap
- Reviewing and refining tracking sources
- Correcting account and facility mappings
- Repairing missing CTM integrations
- Removing overlapping conversion paths
- Refining Google Tag Manager triggers
- Confirming calls and forms reached the appropriate GA4 properties
- Confirming primary conversions reached the correct Google Ads accounts
We also worked directly with the CallTrackingMetrics team to review the findings, confirm the architecture, and agree on the proper order of operations.
That collaboration mattered. This wasn’t a standard implementation involving one site and a handful of phone numbers. It was a network of 17 websites, more than 1,000 numbers, multiple levels of care, numerous traffic sources, separate analytics properties, advertising integrations, and downstream lead data.
The system had to work at scale without becoming impossible to understand or maintain.
A More Dependable Foundation for Marketing
The rebuilt system addressed the structural issues that prevented dynamic number insertion and downstream conversion tracking from working as intended.
The client gained:
- Dependable dynamic number insertion
- A reliable basis for call attribution
- Correct facility-level number routing
- Correct tracking number usage by eliminating hard-coded and incorrectly assigned tracking numbers
- Accurate conversion counts by eliminating duplicate conversions
- Restored missing conversion paths into GA4 and Google Ads
- Alignment between CTM, GA4, and Google Ads
- Cleaner inputs for Google Ads optimization
- A clearer map of how lead information moved between platforms
The marketing team could be confident that calls were attributed according to how people actually found the websites.
We Don’t Treat Tracking as a One-Time Project
Conversion tracking tends to fail quietly.
Sites change, numbers get replaced, forms get updated, and someone eventually drops a tracking number onto a page without realizing it belongs to a specific source.
Meanwhile, the dashboards keep populating, so the problem may go unnoticed by busy internal teams for months. The drift is constant, which is why a one-time fix wouldn’t have held.
Since then, we’ve continued to monitor conversion patterns, investigate unexpected changes, validate tracking when websites or campaigns change, and protect the quality of the data used for reporting and optimization.
Make Sure Your Campaigns Are Optimizing Toward the Right Actions
Unreliable conversion tracking can distort reporting, mislead decision-makers, and send the wrong signals to Google Ads. Meanwhile, automated bidding keeps optimizing toward conversions that were never real, but the budget spent was.
At Astute, we’re comfortable stepping into complicated, inherited systems. We map what exists, identify where the data breaks down, work with the right platform partners, and stay with the problem until the setup is dependable — and keep it that way as the marketing evolves.
Talk with Astute about a conversion tracking and analytics audit.
FAQs About CallTrackingMetrics and Conversion Tracking
1. What Is Dynamic Number Insertion?
Dynamic number insertion changes the phone number displayed on a website based on how a visitor arrived.
Someone coming from a Google Ad may see a different tracking number than someone arriving through organic search, social media, or referral traffic. When the person calls, the tracking platform can connect that inquiry to the source and website activity that preceded it.
2. Why Does CallTrackingMetrics Need a Target Number?
A target number is the permanent phone number that exists on a website before dynamic number insertion occurs.
CallTrackingMetrics needs that number to know when and where to display a source-specific tracking number. If the target number is missing or incorrect, CTM will not display the correct number, preventing accurate call attribution.
3. What Happens When a Dynamic Tracking Number Is Hard-Coded on a Website?
A hard-coded tracking number retains the source assigned to it, regardless of how the visitor actually arrived.
If an organic-search tracking number is entered directly on a page, a visitor from social media may call that number and be incorrectly attributed to organic search.
The call may still connect and appear in the reporting, making the problem difficult to notice.
4. What Should a CallTrackingMetrics Setup Audit Include?
A thorough audit should examine the entire path from the website visitor to each reporting, advertising, and downstream lead platform.
That includes website phone numbers, target numbers, DNI rules, tracking number assignments, fail-safe numbers, traffic source mappings, form integrations, GTM triggers, GA4 events, Google Ads conversion actions, account mappings, and primary conversion settings.
The audit should also include end-to-end testing. Calls and forms need to be verified for correct attribution, routing, destination, and counting.
5. What Causes Duplicate Conversions in Google Ads?
Duplicate conversions often occur when the same lead reaches Google Ads through multiple integrations.
A conversion may reach Google Ads through a direct integration, while the same action is also imported from GA4. Overlapping GTM triggers, duplicated tags, and legacy conversion actions can create similar problems.
Each route needs to be mapped so the team understands which system owns the conversion and how it should reach Google Ads.
6. Can Inaccurate Conversion Tracking Affect Google Ads Performance?
Yes. If calls are attributed to the wrong channel or campaign, Google Ads receives an inaccurate picture of what generated the inquiry.
Duplicate conversions can make campaigns appear more effective than they are. Missing or misattributed conversions can cause effective campaigns to be undervalued. Automated bidding may then optimize against incomplete or inflated information.
Cleaner tracking doesn’t guarantee stronger campaign performance, but it gives Google Ads more reliable inputs to work with.
7. Should Calls and Forms Both Be Primary Conversions?
They can be, provided both represent meaningful leads and each action is counted accurately.
A qualified phone call and a completed inquiry form may both warrant primary status. Phone clicks, partial form interactions, or other lower-intent actions may be more useful as secondary observations.
Those roles should be defined intentionally rather than treating every measurable action as equally valuable.
8. Why Don’t CallTrackingMetrics, GA4, and Google Ads Show the Same Totals?
The platforms may report different totals because they use different attribution models, reporting windows, consent rules, processing methods, and conversion definitions.
An exact match isn’t always realistic or necessary.
What matters is understanding why the numbers differ, eliminating preventable discrepancies, and ensuring the conversions used for reporting and optimization follow a clear and intentional path.
9. How Often Should Conversion Tracking Be Audited?
Tracking should be reviewed whenever a website, phone number, form, campaign, analytics property, advertising account, or platform integration changes.
A broader audit should also be part of ongoing analytics and paid media management. Websites evolve, new tags and numbers are added, conversion actions change, and integrations drift over time.
Regular oversight helps identify problems before months of reporting and optimization are built on unreliable data.