6 Conversion Tracking Mistakes (And How to Fix Each One)
6 Conversion Tracking Mistakes (And How to Fix Each One)
Your tracking says it is working. Is it?
Most small businesses that set up conversion tracking think the job is done once the code is installed. The tag is there. GA4 shows numbers. Meta shows numbers.
But numbers are not the same as the truth. A tracking setup can look perfectly fine and still count the wrong things, count the same thing twice, or miss the actions that actually bring you customers.
These are six conversion tracking mistakes we see again and again, including a few we found on our own website. For each one, here is how to spot it and how to fix it.
1. Tracking installed twice
The mistake. GA4 is added through a plugin or a code snippet, and then added again through Google Tag Manager. Both send data. Every visit and every conversion gets counted twice.
How to spot it. Your page views or conversions suddenly look much better than your sales. Or your bounce rate drops close to zero, which almost never happens with real visitors.
The fix. Pick one place for your tracking. If you use Google Tag Manager, remove GA4 from every plugin, theme setting and code snippet. On our own site, we had an old GA4 snippet still running next to Tag Manager. Switching it off was a two-minute fix that made every number after it honest.
2. Tracking the action your customers do not take
The mistake. Only the contact form is tracked. But most of your enquiries come through WhatsApp or a phone call.
How to spot it. Your reports show two enquiries this month. Your phone shows twenty.
The fix. Track the clicks that start a real conversation. In Google Tag Manager, set up a click trigger for links that contain wa.me for WhatsApp, and tel: for phone calls. Send each one to GA4 as its own event. For a lot of Indian small businesses, the WhatsApp click is the most important conversion on the whole website.
3. Events that never become conversions
The mistake. The events are set up and firing. But nobody marked them as key events in GA4. So they sit in your event list, and your conversion reports stay empty.
How to spot it. You can see your event in GA4 under Events, but your key events count is zero.
The fix. In GA4, go to Admin, then Events, find the event, and mark it as a key event. GA4 used to call these “conversions”. The name changed, the job did not. Only mark the actions that really matter, like enquiries, WhatsApp clicks and purchases. If everything is a key event, nothing stands out.
4. Triggers that break when your website changes
The mistake. A tag is set to fire on a specific page address. Then the page gets renamed or moved during a redesign. The tag keeps waiting for a page that no longer exists.
How to spot it. An event that used to show up every week suddenly drops to zero, right around the time you changed your website.
The fix. After any redesign, open Google Tag Manager and check every trigger that uses a page URL. We found one on our own site still pointing at a services page address we had planned but never used. Update the URL, or use a rule that does not depend on the exact address, like the click text or a CSS class.
5. GA4 knows, but your ads do not
The mistake. WhatsApp clicks are tracked in GA4. But the Meta Pixel only fires on the contact form. So when you run Meta ads, Meta has no idea who actually messaged you.
How to spot it. GA4 shows WhatsApp clicks from Facebook and Instagram visitors. Meta Ads Manager shows almost no results for the same campaign.
The fix. Send the same important actions to every platform you advertise on. In Google Tag Manager, add a Meta Pixel event, like Lead or Contact, using the same WhatsApp click trigger you built for GA4. Meta can only find more people like your customers if it can see what your customers do.
We wrote about what happens when this goes wrong in a business that spent money on a boosted post and got zero clients.
6. Counting yourself and the bots
The mistake. Every time you check your own website, it counts as a visit. Bots and data centres show up as visitors too.
How to spot it. Visits from your own city spike on days you were editing the site. Or you see steady traffic from a city you have never marketed to, with zero engagement. On our site, a stream of visits from a data centre in Oregon turned out to be bots, not customers.
The fix. In GA4, define your own IP address as internal traffic, then switch on the internal traffic filter under Data filters. For bots, look for patterns like one location, very short visits and no clicks, and leave them out when you read your reports.
The 10-minute tracking check
You do not need to be technical to run this. Do it after every website change.
- Open your website in Google Tag Manager’s Preview mode.
- Click your WhatsApp button, your call button and submit your form.
- Check that each one shows a tag firing in Preview.
- Open GA4 Realtime and check that the same events appear.
- In GA4, check that each important event is marked as a key event.
- If you run Meta ads, check the same actions show up in Meta Events Manager.
If anything is missing, you have found your leak.
Good tracking is boring. That is the point.
When tracking works, nobody talks about it. The numbers simply match what is happening in your business. You know which channel brought the enquiry, and you stop guessing where to spend.
Most tracking problems are not complicated. They are just unchecked.
Not sure if your tracking is telling you the truth? Message us on WhatsApp and we will tell you honestly what we find. Got questions first? Read our FAQ.
