Why Google Ads and GA4 Never Agree on Your Numbers

Every owner who runs ads eventually calls me with some version of the same complaint. Google Ads says one thing, GA4 says another, and the numbers are not even close. One report claims 40 conversions, the other shows 17. Somebody must be lying, or something must be broken.

Usually, neither is lying and nothing is broken. The two tools are answering different questions with different rulebooks. Once you see the rules, the mismatch stops being mysterious, and you can stop wasting meetings arguing about which report is right.

They do not even agree on what a conversion is

Start here, because it explains half the gap. Google Ads counts a conversion when someone who clicked an ad later takes the action, and it writes that conversion on the date of the click, not the date of the action. If someone clicks on Monday and books on Thursday, Ads puts the conversion on Monday. GA4 works differently. It records the event in the session where it actually happened, on Thursday, and it may credit that booking to whichever channel touched the visit last, which might not be your ad at all. Same booking, two reports, two different days, sometimes two different owners. Neither made an error.

Attribution windows quietly change the math

Both tools only look back so far. Ads might credit a click from 30 days ago, while a GA4 report might be using a different window for the same data. Change either window and the totals change with it, without a single customer doing anything different. On top of that, both systems model some of the data they cannot observe directly, filling gaps left by blocked cookies and lost signals with estimates. Modeled numbers move when the model changes. That is not a flaw in your setup, it is how the tools now work.

Small numbers hit data thresholds

GA4 protects visitor privacy by withholding or blending data when the counts are small. For a local business, counts are often small. So a report may show fewer conversions than actually occurred, or roll sources together, simply because there were not enough events to display safely. Ads has its own version of this. When your real numbers are in the dozens, a handful of withheld or modeled events is a visible percentage, and it can look like a big discrepancy when it is really just rounding at the edges of two different systems.

Clicks are not sessions either

The same thing happens one step earlier in the funnel. Ads counts clicks. GA4 counts sessions. One person can click your ad twice in ten minutes and start one session. Another can click, land, and leave before the analytics tag loads, and that click never becomes a session at all. If those two numbers matched exactly, I would be suspicious. If your gap is enormous, though, say Ads shows hundreds of clicks and GA4 shows a trickle of sessions, that is worth a real look, which is the comparison test I lay out in your conversion tracking is probably broken. Small gaps are normal. Cliffs are symptoms.

So which number do you trust?

Here is the practical answer I give every client. Use Google Ads numbers to make Google Ads decisions, because that is the data the ads system itself is optimizing with, and it is consistent within its own rulebook. Use GA4 to understand what people do on your site overall, across every channel, not just ads. And use your actual business records, your bookings, your invoices, your call log, as the ground truth both reports answer to. If the phone rang 30 times and neither tool says anything like that, the problem is not the reports, it is the tracking underneath them. That is also why I push owners toward GTM explained for business owners, because a clean, testable tag setup is what makes both reports as close to reality as they can get.

If you want, we can sit down with your two reports side by side on a screen share and I will show you exactly where your gap comes from, line by line, as part of my conversion tracking help. Most of the time it is the rulebooks. Sometimes it is a real break worth fixing, and either way you leave knowing which. Fix sessions are $99 for the first hour, and the free 15-minute Marketing Check is a good first look.