Your Tracking Isn’t Failing. It’s Misleading You.

Alina Popa at her desk, surrounded by Google Ads, Google Analytics, Shopify and code panels connected by tangled lines

Most tracking setups don’t break.
They don’t throw errors.
They don’t scream for help.

They quietly tell you everything is fine.

And that’s the real problem.

When “working” is the most dangerous state

If your tracking was obviously broken, you’d fix it.

No data? Wrong numbers? Zero conversions?
Great. You know there’s a problem.

But when everything looks fine (events firing, conversions showing, reports populating), that’s when decisions start going wrong.

Budgets get increased. Campaigns get scaled. Strategies get locked in.
All based on data that feels trustworthy… but isn’t.

What misleading tracking actually looks like

Misleading tracking doesn’t mean “nothing is tracked”. It usually means:

  • Conversions fire, but not when the real decision happens
  • One user action triggers multiple conversions
  • Conversion value exists, but has no business logic behind it
  • Attribution looks clean, but journeys are fragmented
  • GA4 and Google Ads agree, and they’re both wrong

Nothing looks broken. But everything is slightly off.
And “slightly off” at scale becomes expensive.

Example 1: the purchase that happens twice

A customer buys something, lands on the thank-you page and the purchase tag fires. So far, so good. Then they refresh the page, or come back to it later from their browser history or the order confirmation email. If the tag fires on every page view of the thank-you page, the same order is now counted twice, with its full value.

Every report shows a healthy, “working” purchase conversion. Smart Bidding sees more revenue than you actually made and bids accordingly.

The fix is usually simple: send the order ID as a transaction ID, so both Google Ads and GA4 can recognise and drop the duplicates. The hard part is noticing you need it.

Example 2: the lead that never arrived

A lead form conversion is set up on the click of the “Submit” button. It fires every time someone clicks, including when the form fails validation, when a required field is empty, or when the user clicks three times because nothing seemed to happen.

The conversion count looks great. The CRM tells a very different story.

Example 3: the match quality score that looks great

You send purchases to Meta through the Conversions API, and Events Manager shows a solid Event Match Quality score. Email, phone, IP address, user agent: all there, all hashed, all green.

But that score tells you how well the customer information you send can be matched to Meta accounts. It doesn’t check whether that information belongs to the person who actually bought.

If the email field is filled with the wrong value (the store’s own notification address, a test email left over from the setup, or an email pulled from the wrong field), the parameter is still present, correctly formatted and hashed. The score can stay high. Your purchases are just being matched to the wrong person, or to no one at all.

The only way to catch it is to look at the actual value before it’s hashed. Place a test order, then check in your server container’s preview (or in Meta’s Test Events) that the email being sent is the one you typed at checkout.

Why platforms won’t save you

Google Ads won’t warn you.
GA4 won’t flag it.
Green checkmarks will happily approve everything.

Because platforms can only verify technical execution, not meaning.

They can tell you:

  • the tag fired
  • the event exists
  • the conversion was received

They cannot tell you:

  • if this conversion should exist
  • if it fired at the right moment
  • if it represents real business value
  • if it’s inflating performance

That part is on you.

What I actually check (before trusting any report)

Before I trust any conversion data, I look for answers to questions like these, and I don’t take the answer from a report. I go and test it.

  1. What exact user action creates this conversion?
    I complete the action myself with GTM Preview mode open and check which trigger fires, and when.
  2. Can the same user trigger it more than once?
    Refresh the thank-you page, use the back button, submit the form twice, then count the events in GA4 DebugView.
  3. Is it tied to intent or just interaction?
    A click on “Submit” is an interaction. A lead that reaches the CRM is intent.
  4. What happens if JavaScript fires twice?
    Look for duplicate tags, the same tag in two containers, or a hardcoded tag next to a GTM one.
  5. Does this conversion still make sense if attribution disappears?
    Compare the totals with your source of truth (orders in the shop, leads in the CRM) for the same period, before looking at any channel split.

If those answers aren’t crystal clear, the numbers don’t matter.

The uncomfortable truth

Most accounts don’t need more tracking.
They need less noise and more intent.

More events won’t fix confusion.
More dashboards won’t fix wrong logic.
More automation won’t fix weak foundations.

Clarity comes from restraint, not volume.

Final thought

If your tracking looks correct, don’t relax.
Pause. Question it.

Because the setups that hurt businesses the most aren’t the broken ones: they’re the convincing ones.


If you’re questioning whether your conversion data reflects what’s actually happening in your business, this is exactly the kind of problem we investigate at Ignite Marketing.

Book a call and let’s take a closer look.

Stay curious.

Found this useful? Share it.

Similar Posts