The consent banner
GA4 sets cookies, so in much of the world it needs consent first. Every visitor who declines is a visitor GA4 sees less of, and Consent Mode fills part of the gap with modelling.
GA4 is strong at attribution: which campaign brought people in. Most teams that look elsewhere aren’t unhappy about attribution. It’s the cookie banner, the Exploration needed for every question, the detail that expires, and the fact that it can’t tell them why checkout leaks. Here’s an honest comparison, including when to keep GA4.
| Evidence | Since v4.12 |
|---|---|
| Biggest leak | Delivery → Payment, down 14 points |
| Top friction | Dead click on button "Apply coupon" |
| Who | Keyboard users: 38% succeed against 71% for people using none |
GA4 would show the drop. This shows the button, the release and the people it stopped.
GA4 sets cookies, so in much of the world it needs consent first. Every visitor who declines is a visitor GA4 sees less of, and Consent Mode fills part of the gap with modelling.
Standard reports cover traffic and campaigns. Anything more specific, like a funnel for your actual checkout, usually means building a custom Exploration.
Explorations only reach back as far as the property’s retention setting: two months by default and 14 at most on standard properties. Aggregate reports keep going.
With Google signals or demographics involved, GA4 can withhold rows with small numbers to protect identity. Useful for privacy, frustrating when the small group is the one you need.
Its strengths are attribution, audiences and Google Ads. Questions about usability and what went wrong in the product are secondary.
App analytics come through Firebase, configured separately and reported differently from your website.
| Google Analytics 4 | Metrickle | |
|---|---|---|
| Built for | Marketing attribution and advertising | UX research and conversion: why people drop off, and who for |
| Cookies and consent | Cookies by default, so usually behind a consent banner | Optional cookieless mode: nothing stored on the device, raw IPs never stored |
| Sources and campaigns | Deep, with Google Ads and attribution models | Referrers, UTM campaigns and sources, without ad-platform attribution |
| Funnels | Funnel Explorations, built per question | Saved funnels with time per step and the biggest leak named, filterable by anything |
| Why people leave | Mostly inferred | Rage, dead and error clicks, u-turns, form errors, surveys, heatmaps and consented replay |
| Accessibility | Not measured | Every event carries the device’s accessibility settings |
| Releases | Annotations | A verdict on protected flows after every release, and a case with the lost revenue when one breaks |
| Mobile apps | Through Firebase | iOS, Android, Flutter and React Native SDKs in one portfolio with your sites |
| Data history | Explorations limited to 2 or 14 months | 90 days on Free, a year on Growth, longer on higher plans |
| Price | Free (GA4 360 for enterprise) | Free plan, then paid plans by volume. See pricing |
Leaving GA4 isn’t always the right call. Many teams run both.
In cookieless mode, Metrickle writes nothing to the visitor’s device. Visitors are counted on the server with a salted hash that changes every day, and the raw IP address is never stored. That means the numbers include the people who would have declined a banner.
GA4’s legal position in Europe has shifted over the years: several regulators found particular set-ups unlawful in 2022, and the EU–US Data Privacy Framework changed the picture again in 2023. Whatever tool you choose, confirm your obligations with whoever advises you on privacy law.
One tag, with data-cookieless on public pages. It runs alongside GA4 without interfering.
Turn the GA4 key events that matter into Metrickle goals. Most teams need five or fewer.
Join the goals into your checkout or sign-up funnel, and open the friction on its biggest leak.
Metrickle doesn’t import GA4 data. Export what you need, or link GA4 to BigQuery, before it ages out of your retention window.
Yes. It’s a separate script tag and doesn’t interfere with GA4. Many teams keep GA4 for ad attribution and use Metrickle to find out why people drop off.
No, and they shouldn’t be expected to. GA4 sees less of visitors who decline consent, and the two tools define sessions and filter bots differently. Compare trends, not totals.
Not today. Export the reports you need from GA4, or link it to BigQuery, before your retention window removes the detail.
Yes. Referrers, UTM source, medium and campaign are recorded on every visit, and any funnel or report can be filtered by them. What it doesn’t do is ad-platform attribution modelling.
Not for Metrickle in cookieless mode, since nothing is stored on the device. Whether you need one overall depends on the rest of your site and where you operate. Confirm that with whoever advises you.
Yes. The Free plan covers three apps and 100,000 events a month, with funnels, friction, the accessibility split and one protected flow. See pricing.
Funnels, friction and accessibility on web and mobile, without a cookie banner in cookieless mode. Free to start.
Start free