In brief: every optimisation is only as good as the number it rests on. In nine out of ten accounts we open, at least one central event is measured incorrectly – counted twice, without a value, without consent, or not at all. The tracking audit checks GA4, Google Ads and Meta against each other and tells you which number holds up.
Optimising While Flying Blind
Hardly any company still advertises without an automated bidding strategy. Smart Bidding at Google, Advantage+ at Meta – both learn exclusively from what you report back to them as success. That turns measurement into the actual control desk: send the wrong signal and you do not end up with bad ads, you end up with a system that reliably optimises in the wrong direction.
The insidious part is that nothing looks broken. The reports are full, the curves are climbing, cost per conversion looks respectable. Only when you lay the systems' numbers side by side does it become obvious that GA4, Google Ads and Meta are telling three different truths – and that nobody in the company can say which one is right.
We see it regularly: an account with well-maintained campaigns, a committed team, a decent budget – and a purchase event that has been firing twice ever since the last website relaunch. Six months of work built on a number that was out by a factor of two.
The Seven Most Common Measurement Errors
- 1. The same conversion is counted twice. One tag in Google Tag Manager, a second one hard-coded in the page source. Or a thank-you page view that the visitor repeats simply by reloading. The result is a doubling that runs through every report – and makes automated bidding systematically too generous.
- 2. The purchase arrives without a value. A purchase event without a value is useless to Smart Bidding: the system can no longer tell an 80 € order from an 8,000 € order and optimises for volume rather than revenue. The most common cause is variables that look correct inside the tag but stay empty at runtime.
- 3. Consent Mode v2 does not pass the signal through. Ever since consent became technically mandatory in the tagging setup, it has been the most common reason for sudden measurement gaps: a badly wired banner blocks the tags permanently – or, worse, does not block them at all and you measure without a legal basis. Both are expensive; the second is risky on top.
- 4. GA4 and the ad accounts do not count the same thing. Different attribution windows, different models, different definitions of a "conversion". A gap is normal – you simply need to know how wide it is allowed to be. When Google Ads deviates from GA4 by more than a quarter over a sustained period, there is almost always a genuine error behind it rather than methodology.
- 5. Meta measures without the Conversions API. The browser pixel on its own loses a considerable share of events to iOS settings and ad blockers. Without a server-side return channel, Meta is missing exactly the data points its automation is meant to learn from – and your campaigns look weaker than they are.
- 6. Test data sits in the live data. Internal IP addresses not excluded, test orders not flagged, agency and developer visits counted along with everyone else. In smaller accounts this shifts the conversion rate noticeably – and nobody goes looking for the error there, because it does not look like one.
- 7. The offline close is missing entirely. If your sales team only closes the deal on the phone, that revenue never finds its way back into the ad account. The automation then optimises for form submissions – including the poor ones. Enhanced Conversions and the offline return channel close exactly that gap.
What the Tracking Audit Delivers
We examine every event individually: trigger, condition, value, duplicate. On top of that comes a full GA4 check – data stream, key events, attribution, internal traffic, referral exclusions, data retention – plus the conversion import into Google Ads including Enhanced Conversions, the Meta pixel and Conversions API, and the consent flow from banner through to tag. At the end stands the question that decides everything: does GA4 count the same thing as your ad accounts?
The outcome is a 15–20-page PDF report with a prioritised action plan: one finding and one recommendation per line, sorted by impact – not by how little work it means for us. Not an automatically generated standard score, but a measurement review checked by hand by specialists, one you can start working through on Monday morning.
We also tell you what is not worth doing. For many mid-sized companies, a full server-side setup is the right answer to a question they do not yet have – and it mostly costs time. What you can realistically win back, and from what order of magnitude the effort pays off, is in the report as well.
No Risk: The Results Guarantee
We share the risk: implement at least 50% of our recommendations within six weeks and still see no measurable gain in effectiveness, and you get your money back. If you have questions about the implementation, we support you free of charge by email.
Instead of 499 €, the tracking audit currently costs 299 € net (plus VAT) – until 30/09/2026. For it we need read access to GA4 and Google Tag Manager, plus Google Ads and Meta Business Manager where they exist. We change nothing without your approval.
Want to know whether your numbers hold up?
We check your measurement along a fixed methodology – GA4, Google Ads, Meta and consent – and supply the repair instructions with it. Including a prioritised action plan and results guarantee.
Grünberg.Digital. – AI marketing for SMEs. Sold and processed by Salon Digital Media GmbH.
