Google Ads says 64 conversions. GA4 says 47. Before you rebuild the tracking, understand this: the two platforms are not designed to produce the same number.
They answer different questions, apply different attribution logic and can put the same conversion into different reporting dates.
A difference is normal. A sudden, unexplained change in the difference is what deserves investigation.
Why do GA4 and Google Ads show different conversion numbers?
Google Ads is primarily interested in the contribution made by Google advertising. GA4 looks across the wider customer journey. That difference alone means the totals can diverge even when both implementations are working.
The six checks to make before calling the tracking broken
- 1Check you’re comparing the same action.A purchase in one platform and a thank-you-page view in the other are not equivalent.
- 2Check attribution and reporting dates.An ad click can happen days before the eventual conversion, and the platforms can report that journey differently.
- 3Check conversion windows and counting.One-versus-every settings and different lookback windows change totals.
- 4Check time zones.Different account and property time zones can move conversions across daily and monthly boundaries.
- 5Check consent and lost signals.Consent choices, browser restrictions and modelling affect what each platform can observe.
- 6Check for duplicate or missing firing.A tag firing twice, a missing event, or two primary actions representing the same outcome can create a genuine fault.
Which number should you trust?
Trust the number appropriate to the decision. Google Ads data is useful for understanding and optimising Google advertising. GA4 is useful for cross-channel behaviour and analysis. Your CRM, booking system or ecommerce platform should normally remain the commercial source of truth for actual leads, orders and revenue.
The mistake is asking one dashboard to answer every question.
When should you worry about the gap?
Don’t use an arbitrary percentage as a universal rule. Investigate when the relationship changes materially without a known reason: after a site release, consent change, tag migration, new checkout, new conversion action or campaign restructure.
Also investigate when one system records nothing, totals suddenly multiply, or reported conversions exceed what the underlying business system says could have happened.
Don’t optimise paid media to a broken definition of success
Accurate firing is only half the problem. If the platform is told that every form submission is equally valuable, it will optimise towards form submissions. That can produce an account full of green metrics and a sales team full of poor leads.
That’s the next diagnostic in this cluster: Google Ads generating leads but no sales: what to check.
Tracking problems often look like website problems
If analytics shows traffic but no enquiries, first establish whether the enquiries genuinely aren’t happening. A broken form event and a broken form can look remarkably similar in a report. Our website traffic but no enquiries checklist starts with that distinction.
And if the manual work of reconciling dashboards is itself the problem, see how to identify business processes worth automating.
Frequently asked questions
Should GA4 and Google Ads conversions match exactly?
No. Exact agreement is not the goal because the systems can apply different attribution, timing and counting rules.
Can consent cause Google Ads and GA4 to disagree?
Yes. Consent configuration and the signals available to each product can affect observed and modelled activity. A change in consent implementation is therefore worth checking when a discrepancy suddenly changes.
What should be the source of truth for revenue?
For actual booked revenue, use the transactional system that records the sale: for example your ecommerce platform, finance system, CRM or booking platform. Analytics platforms are attribution and decision tools, not the ledger.
Don’t force the dashboards to agree
If you don’t trust your marketing numbers, we’ll trace the measurement from the commercial outcome backwards and tell you what is structural, what is broken and what is worth fixing.