Comparing four privacy-focused google analytics alternatives

Most users don’t like being tracked as they go around the web. Yet, site owners often have compelling reasons to want to know how well visitors are hopping from one page to the next. This is often attributed to malice, but there are many cases in which there is someone on the end actually wanting to make your life easier. Reducing 404 pages, or identifying pages that are incredibly frustrating, are two solid use cases for wanting to see how traffic is flowing. The tricky part is balancing that need with respect for users’ privacy. Google Analytics is incredibly popular, but integrates analytics with a whole suite of additional objectives, tracking targeted ad spend, predicting demographics of the users, and overall following users around the web as they navigate between sites.

The Austrian Data Protection Authority ruled in 2022 that transferring data to Google through analytics violated GDPR for European users. Collecting analytics, and the company that is selected to collect them on your behalf, matters. The good news is that there are alternatives that both respect users’ privacy, and enable site owners to get insights into how their sites are being used.

Overview of Alternatives

Here I will go over four alternatives to Google Analytics: Plausible, Fathom, Rybbit, and Umami. While these are similar in many ways, there are subtle differences that should be considered depending on your unique situations. A couple of disclosures, I have used Plausible and Fathom, but I have not used Rybbit or Umami. There are no affiliate codes or links in this post. Everything stated is accurate to the best of my knowledge at the time of publication. If you work for one of these companies and I have made an error, please contact me.

Provider	Lowest Plan Cost	Hosting Location	Open Source	Privacy
Plausible	$9 USD / 10k pageviews	EU	6 month delay	Strong by default
Fathom	$15 USD / 1M data points	EU & US	No	Strong by default
Rybbit	$16 USD / 100k events	Unknown	Fully	Requires configuration
Umami	Free / 100k events	Unknown	Fully	Requires configuration

Plausible

Plausible, an Estonian company, contains strong privacy by default and routes all traffic through servers in the European Union. Having been previously fully open-source on the base product, they have since moved to a community edition model that is kept behind the cloud offering on some features. It is very easy to self-host, and uses a clickhouse database. The script is lightweight, coming in at around 3kb.

By default, Plausible uses the IP address and the User agent, along with a salt, to create a unique id for a given visitor. The salt is rotated every 24 hours, ensuring that no consistent data about users can be obtained from the database.

Fathom

Fathom, a Canadian company, also contains strong privacy by default. It selectively routes European traffic to European servers, while sending other traffic to AWS. Fathom Lite is open source, however the current core product cannot be self-hosted and is not open source.

Like Plausible, Fathom uses a daily rotating salt to protect anonymity. Fathom is the second-most expensive offering, however they are an established company and this may be the cheapest if you expect high volumes of traffic. They also include API access in their cheapest plan.

Rybbit

An incredibly new company, with the first commit in January 2025. The main features that differentiate Rybbit from the rest is the dashboard (shows trends, heatmaps and a very nice map view). There are no details that I was able to find about where the company’s servers are located, however pinging the demo server routed to a Hetzner server in Oregon for me.

By default Rybbit does not use a salt for the IP + User Agent hash. I would advise enabling User ID Salting to bring the privacy on par with Fathom and Plausible.

Rybbit is entirely open-source and can be self-hosted with a clickhouse database.

Umami

The biggest standout about Umami is the free tier with a generous event count. It is open-source, easy to self-host, and is available on pikapods. I could not find any details about where the company is located.

For privacy, the client is responsible for determining the id. This means that you could use something completely random like a UUID, or attempt to fingerprint the user using their browser. I wouldn’t advise fingerprinting, and please do not use any user information, such as email, as the id. Instead, use something privacy-preserving and random.

Conclusion

Running an analytics company costs money, especially when they don’t make revenue from user data. If you are considering adding analytics to either a site that you are using for business, or are looking to use it to satisfy your curiosity about your blog, hopefully this post will help highlight that there are options available, and the similarities and differences between them.