GA4 vs Matomo vs Umami: how much analytics does a small business actually need?

A small business can install an analytics platform capable of recording thousands of events, building audiences, analysing funnels, attributing campaigns and segmenting visitors in dozens of ways. That does not mean anybody will ever use most of it.

Before choosing between GA4, Matomo and Umami, I would ask a simpler question: Who is actually going to look at this data, and what decisions are they expected to make with it?

If nobody has an answer, a complicated analytics implementation may create more privacy, maintenance and reporting work without helping the business very much.

The short answer

Choose GA4 when Google Ads, advertising audiences, conversion imports and Google’s marketing ecosystem are important to the business.

Choose Matomo when you want comprehensive analytics but greater control over hosting, data storage and governance.

Choose Umami when you mainly need clear website, campaign and conversion reporting without the complexity of a full marketing analytics platform.

Choose less tracking altogether if nobody in the company has the time or responsibility to use the information you are collecting.

That last option is underrated.

Start with what somebody will actually do with the report

Imagine a small professional-services company. Good questions to ask might be:

Which pages generate enquiries?
Which marketing channels bring qualified leads?
Which campaigns are worth continuing?
Where do people abandon an important form?
Is organic search bringing more useful visitors than last quarter?

That is a perfectly legitimate analytics setup. You probably do not need hundreds of custom events, elaborate visitor segments and seventeen dashboards to answer those questions.

An ecommerce business may need considerably more:

product discovery;
cart behaviour;
checkout steps;
purchases;
revenue;
campaign attribution;
repeat customers;
advertising audiences.

The right amount of analytics depends on the decisions the organisation is capable of making with it.

More data is not automatically better measurement

There is a tendency to treat analytics implementation as a collection exercise. Track every button. Track every scroll depth. Track every form interaction. Track every navigation state. That is easy to do. The harder question is what happens next.

If nobody reviews the data, nobody owns the reports and nobody changes anything based on the result, most of that collection is just technical overhead. It also creates more things to maintain:

event definitions;
tags;
consent rules;
dashboards;
documentation;
access permissions;
data retention;
integrations.

For a small company, a simpler setup that answers five important questions reliably is often better than a sophisticated setup nobody understands.

GA4: strongest when Google Ads matters

GA4 makes the most sense to me when a business is already heavily invested in Google’s advertising ecosystem. A linked GA4 and Google Ads setup can share key events, conversions, audiences and campaign data between the products. That can be valuable if you are actively running campaigns and somebody is using the data to adjust bidding, audiences, landing pages or campaign spend.

GA4 is also extremely flexible. You can build custom events, audiences, explorations, eCommerce tracking and detailed acquisition reporting. That flexibility is useful for an organisation with somebody who knows what to do with it. It can be excessive for a business whose monthly analytics review consists of:

How many people visited the website?

and:

How many contact forms did we get?

In that situation, a lot of the complexity provides little practical benefit.

There is also a data-governance trade-off with GA4

GA4 is a Google service. If your business already depends on Google Ads, that integration can be precisely why you want it. But if you do not use Google’s advertising ecosystem and only need basic website measurement, it is reasonable to ask why Google needs to be part of the setup at all.

That does not mean GA4 is automatically inappropriate. It means there should be a reason for introducing a large third-party analytics platform and the privacy, consent and governance work that comes with it. If a company collects detailed behavioural data through GA4 but never uses that data to make decisions, the company is carrying most of the implementation burden while getting very little value back.

Matomo: more control without giving up much analytical depth

Matomo sits closer to GA4 in terms of analytical ambition. It can support detailed visitor and acquisition analysis while giving the organisation substantially more control over where the data lives. With Matomo On-Premise, the software and database run on infrastructure you control. Matomo explicitly provides access to the stored data and leaves hosting location under your control.

That can be attractive when data sovereignty, internal governance or reducing dependency on advertising platforms matters. But self-hosting is not free in the operational sense. Someone still needs to take responsibility for:

updates;
database maintenance;
backups;
security;
user access;
server resources;
retention;
monitoring.

Matomo Cloud removes much of that infrastructure work, but then you are once again using a hosted provider rather than operating the whole system yourself. Matomo currently hosts its Cloud offering in Frankfurt. So “self-hosted” should be a conscious operating decision, not just a privacy checkbox.

Umami: often enough for a small business

Umami is the option I think is frequently overlooked. It is much simpler than GA4 or a full Matomo deployment, but it no longer means settling for nothing more than page-view counts. Umami currently supports traffic sources, campaigns, UTM tracking, custom events, goals, funnels, journeys, attribution and revenue reporting. For many small businesses, that already covers the questions that actually matter:

Where did visitors come from?
Which pages did they visit?
Which campaigns worked?
Did they complete the important action?
Where did they drop out?
How much revenue can we associate with a campaign?

It is also cookie-free by default and does not automatically collect personal data, according to Umami’s documentation. It can be hosted by Umami or self-hosted on your own infrastructure. That makes it particularly attractive when the business wants useful first-party analytics without building its marketing operation around Google.

The simpler interface can actually be an advantage

This is where I think analytics discussions sometimes miss the organisational reality. A tool with fewer reports can be better if people actually understand those reports. If the owner of a ten-person company opens Umami once a week and understands:

Organic search generated 12 enquiries.

This campaign generated three.

This service page converts much better than that one.

then the analytics system is doing useful work. A GA4 implementation containing vastly more information is not automatically superior if nobody opens it except the agency that installed it. The value of analytics comes from the decisions it changes.

Who owns analytics inside the company?

Before I build a comprehensive measurement setup, I would want to know who is responsible for it afterwards. That person does not need to be a dedicated analyst. It could be:

the owner;
a marketing manager;
an ecommerce manager;
an external consultant;
somebody responsible for paid acquisition.

But somebody should have a reason to look. Otherwise you tend to end up with beautifully implemented tracking that accumulates data indefinitely and produces little more than a monthly screenshot. At that point, simplifying the setup is often more useful than adding another dashboard.

Your operational system is still the source of truth

Whichever analytics tool you choose, it should not become the authoritative record of the business. For leads, that is usually the CRM. For purchases, it is the eCommerce or payment system. For bookings, it is the booking platform.

Analytics explains what happened around those outcomes. It can tell you where visitors came from, what they did before converting and which pages or campaigns appear to contribute. But if GA4 says you sold 97 orders and WooCommerce says you sold 103, the answer to:

How many orders did we actually receive?

is still 103.

This distinction becomes especially important once consent, ad blockers, attribution and modelling enter the picture.

Privacy does not disappear with Matomo or Umami

There is another trap worth avoiding. Self-hosting does not automatically make an analytics setup compliant. Neither does being cookieless. The relevant questions still include:

what information is collected;
whether it constitutes personal data;
why it is processed;
where it is stored;
how long it is retained;
who can access it;
whether information is stored on or read from the visitor’s device;
and what consent or other requirements apply.

Matomo itself explicitly tells users to determine whether consent is required based on their particular configuration and applicable law.

Umami’s default design avoids cookies and automatic collection of personal information, which can make the privacy situation considerably simpler, but the surrounding website and any custom events still matter.

I would therefore choose the analytics platform after deciding what information the business actually needs. Not the other way around.

A practical way to choose

I would start with the organisation rather than the software.

Choose GA4 when

Google Ads is important, you need tight integration with Google’s advertising products, and somebody will actually use the additional campaign, audience and behavioural data.

Choose Matomo when

You need relatively comprehensive analytics but want more control over hosting, retention and access to the underlying data. It is particularly suitable when the organisation has a real reason to own and operate the analytics infrastructure.

Choose Umami when

You want understandable first-party website analytics, campaigns and conversion measurement without running a much larger analytics stack. For many smaller businesses, this is enough.

Keep the setup extremely small when

Nobody has time to analyse visitor behaviour in depth. Track the outcomes that matter, their main acquisition sources and perhaps a few important journey steps. You can always add more later when there is somebody asking questions that require more data.

My preferred starting point

For a small business, I would usually define three layers first.

Traffic

Where are useful visitors coming from?

Important actions

Did they enquire, book, register or purchase?

Business result

Did that enquiry become a customer? Did the order actually complete? How much revenue was generated? Then I would choose the smallest analytics setup that can connect those layers well enough for the people using it.

Sometimes that is GA4. Sometimes it is Matomo. Increasingly, Umami is a perfectly credible answer as well. And sometimes the sensible answer is simply less analytics.

Collecting more customer behaviour only makes sense when the business has both a legitimate reason to collect it and the resources to turn it into something useful. For analytics and measurement projects, I can help define that scope as well as implement it. That includes GA4, Matomo and Umami, consent-aware tracking, conversion measurement and connecting website analytics with the systems that record the actual business result.