Phone click tracking that counts the wrong clicks

Phone call tracking in Google Ads often counts more than phone calls. On one local service account we manage, the old tracking rule treated any event with "phone" in its label as a call, and about one in six of the "calls" in a September fortnight were people clicking into the phone number box of the contact form. Nobody rang anyone. If your Ads report says you got 100 calls and your phone log disagrees, the rule behind the conversion is the first place to look.

What the old rule counted

The account belongs to a local appliance repair business. Before 1 October 2026 its website sent phone events through a tag manager bridge. That bridge fired a "phone call" event for the proper phone click event, and also for any event whose label contained the word "phone".

For 14 to 30 September the site recorded 130 of these "phone call" events. We split them by where they came from:

• Real taps on a phone number link: 108 events from 98 distinct sessions. Almost all were on mobile (106 events).
• Focus on the phone field of the contact form: 22 events from 21 sessions. The form tracked someone clicking into the field with a label that contained "phone", so the bridge counted it.

The second group is not a call. Someone who clicks into the phone number box may then type a number and send a form, which is a lead worth counting once as a form submission. It is not a phone call, and it should not sit in the same bucket.

What changed on 1 October

As part of a site rebuild the bridge was replaced with a single click listener. It records a phone click on every link that starts with tel: and nothing else, and the Ads conversion "Website - phone number click" fires from the same handler. The form field events were removed on purpose.

On the same day the number of recorded phone clicks fell sharply. That looked like a tracking regression, so we checked rather than assumed.

How we checked

We ran three checks. None of them changed anything in the Ads account.

1. Do the buttons fire? We used a scripted browser on a desktop and a mobile viewport, clicked the first visible phone link on two page types, and watched the network calls. Every phone link on the site is a normal tel: anchor, and the listener fired the site event and the Ads conversion each time.
2. Is the old baseline fair? We recounted September without the form field events. Real phone link sessions averaged 6.1 a day. Our first baseline had been higher, and correcting it mattered, because comparing 1 October onward against an inflated number would have overstated the drop.
3. Did the calls move somewhere else? We checked whether people had shifted to the call button inside the ad. Clicks to call, a Google hosted action, fell from about 4.0 a day in September to 1.4 a day in the first five days of October. Calls from ads, the call reporting action, recorded 2 for the whole 30 days. So the calls had not moved to another route.

After 1 October the site recorded 7 phone link sessions over five days, 1.4 a day, and 8 of the 9 events came from mobile. Every source that was still firing was one that fired before. The result: tracking was working, and the volume really was lower. The 1 October changes also removed some copy and a dead booking link, and the five day window included two weekend days, so we did not claim a cause. We re-read it after a full week.

What to do with your own account

1. List every phone related conversion action in Goals, then Summary. Note which are website events, which are call reporting from the ad, and which are Google hosted.
2. Find the rule behind the website one. Ask whoever built the tracking what exactly fires it. A rule that matches a word in a label, rather than a specific click on a tel: link, will count more than calls.
3. Test on a real phone. Tap each number on your own site and confirm one event arrives, and that tapping into a form box produces none.
4. Split taps from calls. A tap on a number shows intent. A call from the ad that connected for a minimum length is a stronger signal. Keep them as separate actions so one does not hide the other.
5. Compare with your phone log for the same dates. They will not match exactly, since some people tap and do not ring, but a gap of several times is a sign of broken counting.
6. Do not compare periods across a tracking change without recounting the old period under the new rule.

If the numbers feed automatic bidding, this matters more. Bidding learns from what you tell it counts. The related problem of Google created actions is covered in Google Ads says 284 conversions, we counted about 30 real leads, and broken tag firing is in Google Ads tracking breaks without anyone noticing.

Where this sits in Ads management

Checking what a conversion actually counts is the first job on any account we take on, because every bid and budget decision later rests on it. You keep your own Google Ads account, add us as a manager and pay Google directly for clicks. Our Ads management add-on is $549 a month in AUD, month to month, and it sits on top of the $298 a month base plan. Details are on the pricing page, or you can get started and tell us what your account reports today.