Comparison

Metrimato vs. Google Analytics

Google Analytics 4 is a good tool for the job it was built for. This page is about the trade-offs that come with using it on an ordinary website in the EU: consent requirements, measurement gaps and, in eligible configurations, modelled data.

Feature	Metrimato	Google Analytics 4
Cookies	None	Yes, first-party cookies Google, Cookie usage on websites
Cookie banner	Not needed	Consent may be required, depending on the implementation and applicable law Google, About consent mode
Personal data	Not collected	Processes device, location and other data, depending on configuration Google, Data collection Google, EU, Switzerland, or UK-focused data and privacy
Data location	Finland	Any country in which Google or its subprocessors maintain facilities Google Ads Data Processing Terms, data transfers
Multi-site overview	Built in	One property at a time in GA4 Standard, roll-up properties are a 360 feature Google, About roll-up properties
Getting started	One line or the WordPress plugin	Property, data stream and tags Google, Set up Analytics for a website and/or app
The figures	Counted, not sampled	Sampling, thresholds and modelling Google, About data sampling Google, About data thresholds Google, Behavioral modeling for consent mode
Script size	2.5 kB	approximately 150 kB gtag.js, compressed transfer size, measured 30 August 2026
Data retention	You choose, up to 24 months (36 with Pro)	User and event level 2 or 14 months Google, Data retention
Getting the data out	CSV and JSON in every tier, a read API with Pro	BigQuery export, which does not include modelled data Google, Behavioral modeling: unsupported features
Data ownership	The customer	Yours, but data-sharing settings can let Google use it as well Google, Data sharing settings
Support	Direct support in Finnish or English	GA4 Standard: documentation and community support

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

This is the part comparisons often get wrong, so it is worth being careful. In basic consent mode, a visitor who does not consent is not measured by GA4. With advanced consent mode, cookieless signals can be sent and eligible properties can use behavioural modelling to estimate missing behaviour at an aggregate level.

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 %: was the average consent rate on sites with a compliant consent design. Sites using nudging averaged about 14 percentage points higher.; etracker Consent Benchmark 2025, a vendor benchmark
20 %: average GA4 underreporting on sites with consent banners, compared with 11 % on sites without one.; 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. We therefore do not extrapolate a Finnish consent rate from studies conducted elsewhere. The consent rates above come from a benchmark published by etracker, a company that sells cookieless analytics, so we name it rather than presenting it as independent research.

What the acceptance rate actually measures

Research on how banners are built explains much 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 websites studied across 31 countries, 67 % showed a consent interface. Only 15 % of those interfaces met the study's minimum compliance criteria, and fewer than half of them (45 %) offered a reject option on the first screen.
So a high acceptance rate is not a clean measure of visitor preference: the design of the banner can materially change the result.

What Google does about the gap

For eligible properties, behavioural modelling can fill part of that measurement gap 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 with analytics storage denied, for at least seven days, and at least 1,000 daily users with consent on seven of the previous 28 days. Sites that do not reach these thresholds are not eligible for behavioural modelling, and Google says meeting them still does not guarantee it.
Modelled figures are not included in the BigQuery export. This means reports containing modelled behavioural data can differ from the corresponding BigQuery export.
In basic consent mode the tags do not load before the visitor answers, so nothing is sent and there is nothing to model in the first place. Advanced consent mode loads them and sends cookieless signals instead.

The other places the numbers shrink

None of this is hidden. It is documented by Google and by the browser vendors themselves, 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 rather than from a third-party analytics domain.
Safari expires the cookie: Safari's tracking prevention removes cookies set in JavaScript and other script-writable storage after seven days without interaction with the site, and caps them to 24 hours when the visit arrived through a link that WebKit classifies as cross-site tracking. This can cause a returning visitor to be assigned a new browser identifier. Metrimato sets no cookie, so nothing expires.
Thresholds withhold data: When a report contains demographics or search terms, Google Analytics may withhold data to prevent individuals being identified. The thresholds are system defined and cannot be adjusted, and Google does not publish the values.
Sampling and the (other) row: Explorations can be sampled when a query exceeds 10 million events in a standard GA4 property. Separately, dimensions with more than 500 distinct values are considered high-cardinality, which increases the likelihood of data being condensed into an (other) row. Metrimato counts the visits and events it receives in full: no sampling, no thresholds, no (other) row.
Retention is short: Google Analytics keeps user and event level data for 2 or 14 months, and that limit shapes explorations and funnel reports rather than the standard aggregate reports. In Metrimato you choose the retention period within your subscription's limits: up to 24 months, and 36 with Pro.
Why the two tools can disagree: When you run both tools side by side, these mechanisms can explain part of the difference between their numbers. Which ones matter depends on your traffic, consent setup, browsers and the GA4 reports you use.

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.

Your advertising depends on Google's conversion data: Metrimato still shows which campaigns brought visitors and which of them reached a goal, because UTM parameters and click IDs come along. What it does not do is send conversion data back to Google Ads. If your advertising strategy depends on Google's automated bidding or conversion optimisation, keep the appropriate Google advertising tags alongside Metrimato.
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 many 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 costs something, and it pays off mainly 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 Pro adds a read API for the 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 measures differently. Metrimato does not use behavioural modelling, and its visit counts are not reduced by analytics consent within Metrimato itself. It does not, however, identify people across days, so unique visitors over a month is a different 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?

They may. If your current analytics setup only measures visitors who consent to analytics storage, Metrimato can record visits that your existing setup does not. The size of the difference depends on your consent setup, audience, browsers and other blocking mechanisms.

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?

Metrimato itself does not set analytics cookies or store an analytics identifier on the visitor's device. If other services on your site require consent, you may still need a consent banner for those services.

Sources

The research findings and third-party figures used on this page are linked to their original studies or supplier documentation whenever available. The links open in a new tab, so you can check them yourself.

Google's help pages change without version numbers, so all of them were checked on 30 August 2026. The script sizes in the comparison table are our own measurement of the compressed transfer size on the same date.

Switch to a measure that respects the visitor

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

I want in! See pricing