Three website configurations missed between 30% and 45% of controlled mobile call tests. Here is the audit your dealership should run before signing another contract.
Your website renewal is 60 days away.
The vendor says, “We track every call.”
We tested seven dealership website configurations for 48 hours. Three failed to record every mobile call correctly. CallRail showed no attributed call for several tests, even though the phone rang and our device log showed the completed call.
One configuration recorded 12 of 20 tests. Another recorded 11. A local custom website recorded 14 because its click event fired inconsistently.
Those test results do not mean each provider misses 30% to 45% of all dealership calls. They show failure rates within our controlled mobile test. When weighted for mobile traffic, call paths and direct-dial behavior, a dealership might underreport 8% to 12% of website-driven calls.
Your dashboard might report 100 calls when the website influenced 112.
We anonymized the providers because results depend on each dealership’s template, tracking configuration, pop-ups, scripts and custom code. This audit measures the seven configurations tested. It does not grade every website operated by a national provider.
The 48-Hour Mobile Call-Tracking Test
| # | Website Configuration | CallRail Test, 20 Calls | Early Pop-Up? | Sticky Call Bar? | PageSpeed Mobile Sample | Failed Tests | Renewal Position |
| 1 | Provider A, enterprise platform | PASS, 20 of 20 | No | Yes | 68 | 0 of 20 | Tracking passed |
| 2 | Provider B, enterprise retail platform | PASS, 20 of 20 | No | Yes | 72 | 0 of 20 | Tracking passed |
| 3 | Provider C, enterprise site with early pop-up | FAIL, 12 of 20 | Yes, at 3 seconds | No | 54 | 8 of 20 | Demand repair and retest |
| 4 | Provider D, CRM-linked website | PASS, 19 of 20 | No | Yes | 61 | 1 of 20 | Review missed call path |
| 5 | Provider E, legacy enterprise template | FAIL, 11 of 20 | Yes, at 2 seconds | No | 48 | 9 of 20 | Demand repair before renewal |
| 6 | Provider F, automotive website specialist | PASS, 20 of 20 | No | Yes | 65 | 0 of 20 | Tracking passed |
| 7 | Provider G, local custom build | FAIL, 14 of 20 | No | No | 42 | 6 of 20 | Require code audit |
Results apply to the configurations tested. PageSpeed scores are single test samples and change with image size, scripts, consent tools, network conditions and third-party widgets.
How We Tested the Seven Website Configurations
We selected seven dealer websites serving the same market. Each site averaged about 500 daily visits during the test period.
Testing ran Friday and Saturday, when mobile shopping and call activity were strongest.
We used:
- CallRail dynamic number insertion
- Call-button click events
- Ten test paths on an iPhone 14
- Ten test paths on an Android device
- Device phone logs
- CallRail call records
- Browser event records
- Google PageSpeed Insights mobile tests
- A five-second pop-up inspection
- A sticky-button scroll test
Each test started in a fresh browser session. We opened a VDP, waited for the displayed number, scrolled through the vehicle photos and tapped the primary call button.
We checked four results for every call:
- Did the button open the native dialer?
- Did the dialer show the assigned tracking number?
- Did the analytics event record the tap?
- Did CallRail assign the completed call to the correct source?
A call button does not pass because the dealership phone rings. The call must reach the store and appear under the correct website source.
Poor attribution changes budget decisions. A website might produce the call, yet the marketing report credits direct traffic, organic search or no source.
Failure #1: The Phone Rang, but the Tracking Record Failed
The three failing configurations used standard telephone links:
<a href=”tel:5551234567″>Call Now</a>
A telephone link opens the phone dialer. The link alone does not guarantee an analytics event or correct source attribution.
Accurate tracking depends on several components working together:
- Dynamic number insertion must replace the store number
- The tracking script must load before the shopper taps
- The button event must fire
- The completed call must reach the assigned number
- CallRail must connect the call with the proper session
Provider C recorded 12 of 20 tests correctly.
Provider E recorded 11.
Provider G recorded 14.
Some missed calls opened before the tracking number replaced the dealership’s main number. Other calls reached the assigned number but failed to record the button event. One custom configuration lost click events after its consent script loaded.
Adding an old onclick command copied from a help page is not enough. Your website provider and call-tracking company need to test the approved event code, number-swap rule and attribution record together.
Require a passing test for every call path:
- Header phone number
- Sticky mobile call button
- VDP call button
- Service call button
- After-hours number
- Pop-up phone link
- Footer phone number
Action: Complete 20 controlled mobile calls. Match every device record with the tracking dashboard.
Failure #2: Early Pop-Ups Blocked the Button and Delayed Number Swaps
Providers C and E displayed lead pop-ups within two or three seconds of page load.
Each pop-up covered most of the mobile screen. The call control sat behind the overlay. The close icon appeared in the top-right corner and measured about 12 pixels on one configuration.
The shopper had to stop, find the icon, close the window and locate the call button again.
The interruption also affected tracking.
The pop-up loaded during the same opening period as the dynamic number-insertion script. Several test calls started before the assigned number replaced the dealership’s base number.
The calls reached the store. CallRail did not assign them to the tested website session.
Your CRM might still contain the prospect. Your website report loses the source, which damages paid-search reporting, VDP analysis and vendor ROI.
Delay promotional overlays for at least 25 seconds. Engagement triggers offer another option, such as opening the message after two VDP scrolls. The overlay should never cover the persistent mobile call bar.
Action: Record the screen from page load through the first call tap. Send the recording and matching call log to your vendor.
Failure #3: Slow Pages Created Untracked Call Paths
The three failing configurations posted PageSpeed mobile samples between 42 and 54. None displayed a persistent bottom call bar.
A low PageSpeed score does not prove call tracking will fail. The problem in our test involved script order and timing.
Large images, chat widgets, consent managers and pop-ups loaded before the tracking code. Shoppers saw a phone control before the tracking number appeared.
One site displayed the store’s regular number for almost three seconds. The assigned tracking number appeared later.
The call control also disappeared after the user scrolled through two screens of vehicle photos.
A buyer ready to call had to scroll back to the top or find another phone number. Some test paths ended at an untracked number in the footer or a search result.
Set these renewal requirements:
- Persistent Call, Text and Directions controls
- Buttons at least 44 pixels tall
- Tracking script ahead of nonessential widgets
- Stable dynamic number after page load
- Median mobile PageSpeed score above 65
- No overlay covering the call bar
Use 65 as a contract target, not a universal quality grade. Test several times on cellular data and use the median score.
Action: Run the call test outside dealership Wi-Fi. A fast internal connection might hide the same delay shoppers experience.
What the Passing Configurations Did Right
Providers A, B and F recorded all 20 calls. Provider D recorded 19 and required a review of one after-hours call path.
The three clean passes shared five traits:
- The tracking number appeared before the call control became active
- The click event fired when the shopper tapped
- The sticky call control stayed visible during scrolling
- No pop-up covered the button during the first five seconds
- Completed calls matched the correct source in CallRail
Their PageSpeed mobile samples ranged from 65 to 72. Faster loading helped, though script order had more influence than the score alone.
Ask your provider to show the live tracking implementation. Then place the calls yourself.
A screenshot of installed code does not prove attribution works.
The Renewal Email to Send 60 Days Before Expiration
Subject: Mobile call-tracking repair required before website renewal
“We completed a 48-hour mobile call audit using 20 controlled calls across iPhone and Android.
Only 12 of 20 calls received correct website attribution. Eight calls reached the store but failed tracking or source assignment. Attached are our device logs, CallRail records and screen captures.
Before renewal, please complete these repairs:
- Repair mobile call-button event tracking.
- Verify dynamic number insertion on every VDP call path.
- Delay the opening pop-up for at least 25 seconds.
- Add a persistent mobile Call, Text and Directions bar.
- Reach a median mobile PageSpeed score above 65 on the tested template.
- Complete a joint 20-call retest with our team.
Please provide a completion date within five business days. Renewal depends on a passing retest.
We also request a $400 monthly service credit for the documented underreporting period, subject to our contract terms.”
Do not accuse the provider of losing sales unless your phone and CRM records prove missed customer connections. The evidence might show misattribution instead.
Misattribution still needs repair. Your dealership uses those reports to renew vendors, move ad money and judge VDP performance.
Vendor Math: 100 Reported Calls Might Mean 112 Website Calls
Assume your reports miss 12 website-driven calls each month.
12 calls × 30% appointment rate = 3.6 appointments
3.6 appointments × 50% show rate = 1.8 shows
1.8 shows × 40% close rate = 0.72 sales
0.72 sales × $2,200 gross = $1,584 per month
$1,584 × 12 months = $19,008 per year
The $19,008 represents misattributed gross, not confirmed lost gross. Your store might still sell those customers. The website receives no credit, which leads management to cut the wrong channel or renew under false performance data.
The 60-Day Website Renewal Audit
Put these seven checks on your calendar before the contract renews:
- Place 20 mobile calls across iPhone and Android
- Match every device call with a CallRail record
- Verify the click event for each button
- Confirm dynamic number insertion before tapping
- Delay pop-ups for at least 25 seconds
- Test the sticky call bar after scrolling
- Record the median PageSpeed mobile score
Your provider’s reporting deck does not settle the call-tracking question.
Twenty completed calls should produce twenty matching records. Anything less requires a repair, another test and written renewal terms.
Related Reading
We Audited 50 Dealer Websites: 5 Mobile Errors Costing 15% of Calls
9 Mobile CTA Tests That Generate 15% More Dealership Calls in 2026
Sponsored by Gas.net — powering dealership growth through intelligent data.
Alt text: “Gas.net connects franchise dealers with integrated analytics and marketing tools.”