GA4 conversion rate: how it's calculated and why it looks wrong

GA4 doesn't have a conversion rate. It has two, and they answer different questions.

Session conversion rate is the percentage of sessions in which at least one conversion happened. User conversion rate is the percentage of users who converted at least once. Same conversions, different denominators, two different numbers on the same screen.

Neither of them will match Google Ads, and that isn't a bug. GA4 and Google Ads count conversions on different models, in different time windows, with different attribution. Expecting them to agree is the mistake.

This article covers exactly how each rate is calculated, why the two GA4 numbers differ from each other, why both differ from Google Ads, and which one you should actually report.

The two conversion rates, precisely

Both are in GA4, both are called "conversion rate," and GA4 doesn't make it obvious which one you're looking at.

Session conversion rate

sessions with at least one conversion / total sessions

If 1,000 sessions produced 30 sessions containing a purchase, that's 3%.

Note the wording: sessions in which a conversion happened, not number of conversions. If one session contains three purchases, it still counts once in the numerator. The rate can never exceed 100%.

User conversion rate

users who converted at least once / total users

If 700 users produced 30 who bought, that's 4.3%.

The user rate is almost always higher, because one user often needs several sessions before converting. Three sessions, one purchase: session rate 33%, user rate 100%.

That gap between the two isn't an error. It's a measure of how much consideration your product needs. A wide gap means people come back repeatedly before buying, which is normal for high-consideration B2B and expensive e-commerce. A narrow gap means people decide in one visit.

Is your conversion tracking even firing correctly? The Free GTM Audit checks the tags feeding these numbers. Enter a URL, get a report in about 10 minutes. No signup.

Why GA4 and Google Ads never agree

This is the question I get more than any other, and the answer disappoints people: they aren't supposed to agree.

Four differences, each one enough on its own to create a gap.

1. Different dates for the same sale

Google Ads assigns a conversion to the date of the click. GA4 assigns it to the date of the conversion.

Someone clicks your ad on Monday and buys on Thursday. Google Ads books that sale on Monday. GA4 books it on Thursday. Look at a single day and the numbers can't match, no matter how perfect your tracking is.

This one difference explains most of the gaps people spend weeks investigating.

2. Different attribution

Google Ads gives credit to the Google Ads click. GA4 runs its own attribution model across all channels and might hand the credit to organic search, email, or direct.

If a user clicks a paid ad, leaves, comes back through organic search, and buys, Google Ads may claim the conversion while GA4 credits organic. Both are internally consistent. They just disagree about who earned it.

Attribution only works if GA4 knows where the traffic came from in the first place. If a large share of your sessions is landing in Unassigned or Direct, the model is assigning credit from incomplete data, and no amount of reconciliation will fix that.

3. Different counting rules

Google Ads lets you choose whether to count every conversion or only one per click. GA4 counts every conversion event it receives.

An e-commerce customer who buys twice in one session generates two conversions in GA4, and possibly one in Google Ads, depending on your setting.

4. Different data reaching each system

Google Ads conversions can fire through the Google Ads tag, enhanced conversions, or offline import. GA4 conversions only exist if the GA4 event arrived and survived consent, ad blockers, and parameter validation.

When a user denies consent, Google Ads may still model the conversion. GA4 may not record a usable session at all. Two systems, two different views of the same customer.

What a healthy gap looks like

In the properties I have audited, a 5-15% difference between GA4 and Google Ads conversions is normal and not worth chasing. Above 20%, something is usually wrong: double-tagged conversions, a misconfigured conversion event, or consent handling that differs between the two.

The trap is treating the gap itself as the problem. The gap is expected. Its size is the signal.

Why your conversion rate suddenly changed

A conversion rate is a fraction. When it moves sharply, the denominator is the more likely culprit, and almost nobody checks it first.

Ewa runs e-commerce at a home goods retailer. Her conversion rate dropped from 2.8% to 1.4% in a week. Nothing had changed in the shop, the ads, or the prices. The team spent nine days investigating checkout friction, pricing, and stock levels.

The actual cause: a developer had added the gtag.js snippet directly to the theme, while GTM was already firing a GA4 configuration tag. Every pageview counted twice. Sessions doubled. Conversions didn't. The rate halved exactly as arithmetic requires.

Sales hadn't fallen at all. The denominator had doubled.

Before you investigate a conversion rate drop, check these three things, in this order:

1. Did sessions jump? If sessions rose sharply while conversions stayed flat, you have a counting problem, not a sales problem. Look for double-tagging first: view source, search for gtag(, and check whether GTM also has a GA4 configuration tag.
2. Did conversions fall? If conversions dropped while sessions stayed flat, check whether the conversion event is still firing at all. A tag that stops firing looks identical to a business that stopped selling.
3. Did someone change the conversion definition? Marking an extra event as a conversion silently changes the numerator for everyone, retroactively in some reports.

Step 1 catches more real incidents than steps 2 and 3 combined. A doubled denominator is the most common cause of a "collapsed" conversion rate I see, and it is the one nobody checks, because the graph looks like a business problem.

The naming change nobody told you about

If your reports stopped saying "conversions" and started saying "key events," nothing broke. Google renamed it.

In GA4, the events you mark as important are now called key events. The word "conversion" is reserved for what Google Ads counts. The idea was to stop two different systems using one word for two different things, which was a real problem worth solving.

In practice it created a transition period where the same number carries different names in different screens, and older documentation, dashboards, and Looker Studio reports still say "conversion."

What this means for you:

• Key event = an important event in GA4. This is what feeds GA4's conversion rate.
• Conversion = what Google Ads counts and bids on.
• A key event in GA4 becomes a conversion in Google Ads only if you mark it as such and the Google Ads link is working.

If your Looker Studio dashboard suddenly shows blanks where conversions used to be, this rename is the first thing to check. The metric did not disappear. Its field name changed, and the dashboard is still asking for the old one.

Which conversion rate should you actually report?

Pick one, write down the definition, and stop switching.

Use session conversion rate when:

• You're optimising a landing page, an ad, or a checkout flow.
• You want to compare traffic sources fairly, since sources bring sessions.
• Your product is bought in a single visit.

Use user conversion rate when:

• Your sales cycle spans multiple visits, which is most B2B.
• You care about how many people eventually buy, not how efficient each visit is.
• You're reporting to a board that thinks in customers, not sessions.

The worst outcome isn't picking the wrong one. It's having three people in the same meeting quoting three different numbers, all of them technically correct, and no one realising they aren't talking about the same thing.

Write the definition into the report itself. One line: "Conversion rate = sessions with a purchase / total sessions, GA4, last non-direct attribution." That single line prevents more arguments than any dashboard.

Frequently asked questions

How is conversion rate calculated in GA4? GA4 calculates two. Session conversion rate is sessions with at least one conversion divided by total sessions. User conversion rate is users who converted at least once divided by total users. The user rate is normally higher, because users often need several sessions to convert.

Why is my GA4 conversion rate different from Google Ads? Because they measure different things. Google Ads books a conversion on the date of the click, GA4 on the date of the conversion. They also use different attribution models and different counting rules. A 5-15% gap is normal.

What is a good conversion rate in GA4? There's no universal benchmark worth trusting, because it depends entirely on the traffic mix and the product. Branded search converts many times better than cold display traffic on the same site. Compare yourself to your own history, segmented by channel, not to an industry average.

Why did my conversion rate suddenly drop by half? Check whether sessions doubled before assuming sales fell. Double-tagging (gtag.js and a GTM GA4 tag firing together) doubles the denominator and halves the rate overnight without a single lost sale.

Can conversion rate be above 100% in GA4? Session and user conversion rates can't, because they count sessions or users that converted at least once. Metrics that divide total conversion events by sessions can exceed 100% if users convert multiple times per session.

Should I use session or user conversion rate? Session rate for optimising pages, ads, and checkout. User rate for multi-visit sales cycles and reporting on how many people eventually buy. Pick one, define it in writing, and stay consistent.

What is the difference between a key event and a conversion in GA4? A key event is an important event inside GA4, and it is what GA4's conversion rate is built on. A conversion is what Google Ads counts and bids on. Google renamed the GA4 side to stop the two systems sharing one word for two different things.

Next steps

Most conversion rate problems in GA4 aren't conversion problems. They are definition problems and denominator problems.

Start by finding out which of the two rates your dashboard is showing. Then check the denominator before you investigate the business: a sharp drop in conversion rate with flat conversions is a tracking incident, not a sales incident, and the fastest test is looking for a doubled session count.

Then stop trying to reconcile GA4 with Google Ads to the decimal. They book conversions on different dates using different attribution. A gap is expected. Only the size of the gap is worth your time.

If you want the tracking layer behind these numbers checked properly, the Free GTM Audit inspects the tags that feed them. And if you would rather someone verify conversion accuracy every month than discover a broken tag by accident, that's what independent GA4 monitoring is for. €150/month, written report, no calls, cancel anytime. Full pricing here.

---

Sources and further reading

• GA4 conversion rate metrics (Google)
• Why Google Ads and GA4 conversions differ (Google)
• GA4 attribution models (Google)