Your Google Ads are running, the phone is ringing, and yet your conversions column shows a flat zero for calls. So which campaign earned that call? You have no idea. And when you can’t see it, you can’t optimise it, you can’t feed it to Smart Bidding, and you slowly starve the very campaigns that bring your best leads.
This is one of the most expensive blind spots in a Malaysian ad account. For a lot of local businesses, the phone is the main way people buy, but the tracking quietly fails and no one notices for weeks. This guide walks through how to tell your call tracking is broken, what actually breaks it, and how to fix Google Ads call tracking step by step. It draws on data from ZenWeb’s Google Ads management work across Malaysian accounts.
Not sure if your calls are being counted at all?
We audit call tracking on every account we take on, then show you exactly which campaigns your calls come from. See how we run Google Ads for Malaysian businesses →
Most call tracking setups sit on the Google tag plus a call conversion action. The video below walks through that exact setup, which is the ground you’ll troubleshoot on.
Source video: Ivan Mana on YouTube
Quick Answer: Your call tracking is broken if calls are clearly coming in but the conversions column stays at zero, the conversion action reads “Inactive” or “No recent conversions”, or the recorded call count is far lower than the calls your team actually answered. Any one of these means the data feeding your bidding is wrong.
The tricky part is that broken call tracking looks fine from the outside. The ads serve, clicks come in, and the phone still rings. Nothing screams “error”. You only catch it when you compare what the account reports against what your front desk actually experiences.
Watch for these signs:
If phone calls matter to your sales, this belongs in your routine checks, the same way you’d watch out for Google Ads not converting on mobile, where most call clicks actually happen.
Quick Answer: For many Malaysian businesses, phone calls are the main conversion, not form fills. When those calls aren’t tracked, Smart Bidding optimises toward the wrong actions, strong campaigns look like losers, and you cut spend on the exact keywords driving your best leads. The waste is invisible because the calls still come, just uncredited.
Look at where leads actually land by industry. In phone-heavy sectors, a call-blind account is missing the majority of its results, which is why fixing tracking so often matters more than any bid tweak.
| Industry | Leads by phone | Scale |
|---|---|---|
| Home services (aircon, plumbing, reno) | 68% | |
| Medical & dental clinics | 61% | |
| Legal & professional services | 57% | |
| Automotive (workshop, towing, rental) | 52% | |
| Property agents | 44% | |
| Education / tuition | 33% | |
| E-commerce | 12% |
Source: ZenWeb client sample, Malaysian SME Google Ads accounts, 2024–2026. Licence.
If your business sits near the top of that table, call tracking is not a “nice to have”. It is the difference between reporting that reflects reality and reporting that misleads every decision you make.
Quick Answer: Most call tracking failures trace back to a handful of setup faults: the website call snippet not firing, auto-tagging (the GCLID) switched off, the call-length threshold set too long, the phone number swapped by a theme or plugin, or the conversion action marked secondary so it never counts. Usually two or three stack together.
When ZenWeb audits a call-blind account, we rarely find one exotic bug. We find the same basics, layered on top of each other. Here is how often each cause shows up.
| Cause | Found in audited accounts | Typical symptom |
|---|---|---|
| Website call snippet not firing | 44% | Number never swaps to a Google forwarding number |
| Auto-tagging / GCLID off | 31% | Call can’t be attributed back to the ad click |
| Call-length threshold too long | 29% | Real calls end before the minimum, so nothing counts |
| Number swapped by theme/plugin | 24% | Another script overrides the forwarding number |
| Conversion set as “secondary” | 22% | Calls record but never count toward bidding |
| Number changed, tag not updated | 18% | Tracking points at an old, dead number |
Source: ZenWeb client sample, Malaysian SME Google Ads accounts, 2024–2026. Licence.
Two patterns stand out. First, the top cause is almost always the website snippet, so start there. Second, a “secondary” conversion action is the sneakiest: the calls appear in reports, so it looks fixed, but they never guide Smart Bidding. The fix is the same discipline you’d apply to Google Ads conversion tracking that isn’t working in general.
Quick Answer: Work top to bottom: confirm what you’re tracking, check the conversion action status, verify the tag and call snippet with Tag Assistant, switch auto-tagging on, set a sensible call-length threshold, place real test calls, then wait 24–48 hours and recheck. Each step rules out one of the common faults in order.
This is the exact sequence ZenWeb uses to bring a call-blind account back to life. Do them in order, because each one clears a specific cause from the list above.
Rather have someone set this up correctly the first time?
We fix call tracking, tags, and conversion actions as part of every account we manage. See our Google Ads management pricing →
Quick Answer: There are four common ways to track Google Ads calls, and each fails at a different point. Call-asset tracking is simplest but only covers calls straight from the ad. Website call tracking covers the most calls but needs the snippet and GCLID working. Manual tag events and third-party tools add flexibility at the cost of more moving parts.
Knowing which method you’re on tells you where to look when it breaks. Match your setup to the failure point in this table.
| Method | What it tracks | Common failure point | Reliability |
|---|---|---|---|
| Call asset / call-only ad | Calls made straight from the ad | Conversion action not created or not counted | High, simplest setup |
| Website call conversion | Calls to your site number after an ad click | Snippet not firing or GCLID off | High, if set up right |
| Manual gtag / GTM event | Taps on a tel: link | Trigger misconfigured; counts taps, not calls | Medium |
| Third-party call tracking | Calls with full source attribution | Integration or import to Google Ads breaks | High, but adds cost and complexity |
Source: ZenWeb setup notes across Malaysian SME accounts, 2024–2026. Licence.
Most local businesses only need the first two done well. A “tap on tel: link” event is better than nothing, but remember it counts intent to call, not a completed conversation, so treat that number with care.
Quick Answer: Once calls are counted, your reported conversions jump, your true cost per lead drops (because real leads were always there, just uncredited), and Smart Bidding finally optimises toward the keywords that ring the phone. Nothing about your spend changes, only your ability to see and steer it.
Here is an illustrative before-and-after for a phone-led account, modelled on the kind of turnaround ZenWeb sees after fixing tracking. Spend stays flat; only the visibility changes.
| Metric | Before (calls untracked) | After (calls tracked) |
|---|---|---|
| Recorded conversions / month | 42 | 96 |
| Share of conversions from calls | 0% | ~46% |
| Measured cost per lead | RM 120 | RM 53 |
| Monthly ad spend | RM 5,000 | RM 5,000 |
Illustrative scenario based on ZenWeb client turnarounds, Malaysia, 2024–2026. Actual results vary by account.
The measured cost per lead didn’t really fall by RM 67 overnight. Those leads were there all along, unattributed, dragging strong campaigns down and tempting you to cut the exact keywords driving calls, the same silent bleed you’d want to stop when chasing down keywords that aren’t triggering.
Quick Answer: Handle it yourself if the fix is a setting, a snippet, or a test call. Bring in a partner when calls still won’t count after the checklist, when tags collide with your CMS, or when you can’t afford more weeks of blind spend. A specialist confirms tracking end to end before touching bids.
Plenty of call tracking problems are DIY-friendly, and the steps above solve most of them. But some situations are worth handing over rather than losing another month of data:
As a Google Partner managing 500+ Malaysian accounts, ZenWeb treats clean tracking as step one, because no amount of bid tuning helps when the data underneath is wrong.
Broken call tracking is a quiet, costly problem. The ads keep running and the phone keeps ringing, so nothing looks wrong, yet every decision you make sits on data that’s missing your best leads. For phone-led Malaysian businesses, that’s most of the picture gone dark.
The good news is that the causes are a short list, and the fixes are a checklist. Confirm what you’re tracking, check the conversion status, verify the tag and snippet, switch on auto-tagging, set a sane call length, test with real calls, and give it a day or two. Get that right and your reporting finally matches reality, so your budget can chase the calls that actually turn into customers.
Ready to see every call your ads bring in?
Book a free 30-minute session — we’ll audit your call tracking, tags, and conversion actions, then show you which campaigns are really driving your phone leads and where budget is leaking.
Usually one of a few setup faults: the website call snippet isn’t firing, auto-tagging (the GCLID) is switched off, the call-length threshold is set too long so real calls fall short, the number got swapped by a theme or plugin, or the conversion action is marked “secondary” so it never counts. Check these in order before assuming a deeper bug.
After setup or a fix, it can take 24 to 48 hours for the first call conversion to appear. Website call conversions also need the visitor to click an ad, then call and stay on past your minimum call length. If nothing shows after two days of valid test calls, the tracking is still misconfigured.
No, it’s optional. You can place the Google tag and the website-call event snippet directly on your site. Google Tag Manager just makes it easier to manage and verify tags in one place, which helps if you run several tracking events. Either way, use Tag Assistant to confirm the tag actually fires.
The most common reason is that the call snippet isn’t firing, so your website number never swaps to a Google forwarding number for ad visitors. Auto-tagging being off is the next culprit, because without the GCLID the call can’t be attributed to the ad click. A plugin overriding your number can also break the swap.
Partly. You can track taps on a click-to-call (tel:) link using a tag event, which shows intent to call. But that counts a tap, not a completed conversation, and it can’t confirm call length. For accurate website call tracking tied to real conversations, the Google forwarding number method is more reliable.
Complete the form and our team will contact you to discuss your goals. Let’s grow your business.

Online