GA4 and Google Ads conversions don’t match: diagnose it by asking

Written by Uptimal, the team that builds and operates Upbuild and runs marketing analytics for multinational clients across the globe. Everything below reflects how the tools actually behave, including the parts the platforms would rather not discuss.

The most reopened ticket in marketing analytics: Google Ads says 412 conversions, GA4 says 288, and a stakeholder wants to know which one is lying. Neither is. They are answering different questions, and the fastest way to see which difference is driving your gap is to query both from one place and compare like for like.

The four causes, ranked by how often they are guilty

Attribution. Google Ads credits its own ads (last Google click by default, or data-driven within Google surfaces). GA4 shares credit across every channel. A campaign that closes what email opened will always look stronger in Ads than in GA4.

Counting time. Google Ads books the conversion on the day of the click; GA4 books it on the day it happened. Any window that cuts across a considered-purchase cycle disagrees by construction, worst at month boundaries.

What gets counted. View-through conversions, "every" versus "one" counting, imported offline events on one side only, and enhanced conversions all pad one column and not the other.

Consent and modeling. After Consent Mode, both systems model unconsented users, differently. The gap is no longer a fixed percentage; it drifts with your consent rate, which is why last quarter’s "known 15% delta" quietly became 24%.

Diagnosing your own account, conversationally

With Google Ads and GA4 connected through one read-only MCP connector, the diagnosis is one conversation in Claude, ChatGPT, Copilot Studio or Gemini Enterprise: same 30 days from both, split by campaign and conversion action, then let the pattern talk. A uniform gap across campaigns points at attribution or consent; a gap concentrated in one conversion action points at counting settings; a gap that grows at month end points at conversion-time lag. Ask for the split by device and by new versus returning and the story usually finishes itself.

What to do with the answer

Do not "fix" the gap; document it. Bid and budget with the platform number, judge the business with GA4, and put the one-paragraph explanation of your specific delta in the report footer. Then set a guard so drift stops surprising you: an alert when the ratio between the two moves more than a few points week over week is the cheapest early warning system in measurement.

Frequently asked questions