Comparison

Metrimato vs. Google Analytics

Google Analytics 4 is a good tool for the job it was built for. This page is about what that job costs on an ordinary website in the EU: a consent banner, a large share of your visitors, and a report that is partly estimated.

Feature	Metrimato	Google Analytics 4
Cookies	None	Yes
Cookie banner	Not needed	Required
Personal data	Not collected	Collected
Data location	Finland	Global
Every site in one view	Yes	One property at a time
Getting started	One line or the WordPress plugin	Property, data stream and tags
The figures	From every visit	Sampling, thresholds and modelling
Script size	2.5 kB	approx. 50 kB
Data retention	You choose, up to three years	User level 2 or 14 months, 14 by default
Getting the data out	CSV, JSON and a read API in every plan	BigQuery export, without the modelled part
Data ownership	The customer	Shared
Support	From a person, in Finnish or English	Help articles and a community forum

The banner does not cost you visitors. It costs you the record of them.

This is the part most comparisons get wrong, so it is worth being careful. Analytics cookies need consent in the EU. A visitor who ignores the banner has not consented, so Google Analytics does not count that visitor. The visitor came, read and left, and the report never knew.

25 %: of everyone who was shown a cookie banner accepted all cookies. A third never touched the banner at all.; Advance Metrics, 2023, international business sites
40 %: consented when the banner had an equally visible reject button. With a nudging design the figure rose to 54 %.; etracker Consent Benchmark 2025, 500 German sites
20 %: of conversions were missing from Google Analytics on sites with a consent banner. Without a banner the gap was 11 %.; Orbit Media, 2024, 60 comparisons across 33 accounts

Read those numbers carefully, because their denominators differ. Consent platform marketing usually quotes a consent rate, which counts only the people who clicked something. The figure that matters for your report is the share of everyone who was shown the banner, and it is considerably lower. There is no published measurement for Finland specifically, and we are not going to invent one. Finnish guidance is on the strict side, which argues for a lower share rather than a higher one.

The acceptance rate measures the banner, not the visitor

Research on how banners are built explains most of the spread between those figures.

Removing the reject button from the first screen raises acceptance by 22 to 23 percentage points.
When nothing is pre-ticked and consent has to be given per category, almost nobody accepts everything: under 0.1 % on mobile and effectively 0 % on desktop, in a field experiment with 82,890 visitors.
Of 254,148 banners studied across 31 countries, 15 % met the minimum legal requirements and only 21 % offered a reject option at all.
So a high acceptance rate usually says something about the design of the banner, not about what the visitor wanted.

What Google does about the gap

Consent mode fills the hole with machine learning. That is a reasonable engineering answer, and it is worth knowing what it means for the number you are reading.

Google's own documentation says the behaviour of visitors who declined is estimated from the behaviour of similar visitors who accepted. It is a model, not a measurement.
Modelling needs at least 1,000 events a day where analytics storage was denied, for at least seven days, and at least 1,000 daily users with consent on seven of the previous 28 days. A smaller site never reaches that and simply loses those visitors.
Modelled figures are not included in the BigQuery export. The report and the raw data disagree by exactly the estimated part.
In basic consent mode nothing is sent at all when consent is denied, so there is nothing to model in the first place.

The other places the numbers shrink

None of this is hidden. All of it is in Google's own documentation, which is why we can be specific about it.

Ad blockers and tracking protection: Blocking lists know Google's tracking script by name. Around three in ten internet users worldwide used an ad blocker in 2025. Metrimato's script is served from our own domain and matches none of the known patterns.
Safari expires the cookie: Safari's tracking prevention expires the cookie Google Analytics sets after seven days, and after 24 hours when the visit arrived through a link it treats as tracking. A returning visitor is counted as new. Metrimato sets no cookie, so nothing expires.
Thresholds hide rows: When a report contains demographics or search terms, Google Analytics hides rows to prevent individuals being identified. Google does not publish the limit and it cannot be adjusted.
Sampling and the other row: Explorations sample above 10 million events, and a dimension with more than 500 values collects the rest into a row called other. Metrimato counts every visit and every event, with no sampling at any size.
Retention is short by default: Google Analytics keeps user-level data for 2 or 14 months, 14 by default, and that limit shapes explorations and funnels rather than the standard reports. In Metrimato you choose the retention period within your plan's limits, up to three years.
Every one of these is a reason the two tools disagree: When you run both side by side, these are the differences you will see, and knowing which direction each one pushes makes the comparison useful rather than alarming.

Honestly

When Google Analytics is the better choice

There are jobs Metrimato cannot do and will not learn to do, because doing them would mean following individual people. If one of these is your job, keep Google Analytics, with our blessing.

You buy Google Ads on a large budget: This only matters once the budget is large enough that Google's automated bidding needs the conversion data fed back into the ad account. A small business loses nothing here: you still see in Metrimato which campaigns brought visitors and which of them reached a goal, because UTM parameters and click ids come along. What is missing is only the automatic conversion feed back to Google, because Metrimato sends nothing to any advertising platform and never will.
You need one person followed across sessions and devices: Cohorts, retention curves and lifetime value all need an identity that survives. Metrimato's identifier is deliberately gone by morning.
You need multi-touch attribution over months: A purchase that follows a campaign click six weeks earlier cannot be joined to it here. For most businesses this is not the question worth paying for: knowing where people arrive from and which content pulls them in is enough, because the sales show up in your own books and in the shop figures anyway. Modelling a long buying journey always costs something, and it pays off only once you actually move large sums between channels on the strength of it.
Your reporting sits on a data warehouse: Metrimato exports CSV and JSON and has a read API for aggregate figures. There is no visitor-level export, because there is no visitor-level data.

Thinking about switching?

Switching is easy. You can start from a clean slate or bring your Google Analytics history along, Metrimato does both. The guide is here.

Read the guide

Frequently asked

Is Metrimato as accurate as Google Analytics?

It is different, and on an EU site it is usually more complete, because nobody is lost to a banner and no figure is modelled. But it does not identify anyone across days, so unique visitors over a month is not the same measurement and should not be compared directly.

Can I run both at the same time?

Yes, and we recommend it. The two scripts do not know about each other, and a fortnight of overlap tells you more about the difference than any comparison table can.

Will my visitor numbers go up when I switch?

On most EU sites, yes. That is not inflation. It is the visitors who were always there and were never recorded.

Do you support Google Ads conversion tracking?

No. If you advertise, keep the advertising tags. Metrimato measures the site, not the ad account.

Can I bring my Google Analytics history along?

Yes. Connect the site to your Google account with read-only access and import up to 730 days, or upload a CSV file. Imported rows are marked, so an import can be undone.

Does Metrimato need a cookie banner of its own?

No. It stores nothing on the visitor's device and collects no personal data, so there is nothing to ask consent for. If other things on your site need consent, the banner stays for those.

Sources

Every figure on this page comes from a named study or from the supplier's own documentation. The links open in a new tab, so you can check them yourself.

Switch to a measure that respects the visitor

We onboard new companies by invitation so we have time to help everyone get started.

Request an invite See pricing