Fathom Analytics

I signed up with Fathom Analytics for a few months to see what “web analytics” were, and how they are gathered.

Because I’m running this blog myself, on a virtual private server, I have access to Apache Web Server log files. I’ve always found the log files confusing. The format of the log files is clear, but the “who is doing that, and why?” isn’t at all clear. Secondarily, I never quite understood how people gathered the information they talk about in web posts about visitors, unique visitors, etc etc. My reading of Apache log files is that there’s about 5 or 10 scraper and bot visits for every real human. My view of bot vs human visits might be skewed by having run a WordPress honey pot for years, and by sending some web spiders down a hole with no bottom.

What you have to do to use Fathom

To use Fathom, you insert a JavaScript one-liner in all your rendered HTML:

<script src="https://cdn.usefathom.com/script.js" data-site="QUACCKQX" defer></script>

You get a web browser “dashboard” to see all the analytics. Fathom did an excellent job with the dashboard. It’s fast, it’s easy, you can understand almost anything at first glance.

What are “web analytics”?

Beyond something called “bounce rate”, web analytics are just counts of web pages visited, along with human understandable things like desktop/laptop/tablet, interpreting the requesting IP address as a country, and give a guess about requesting operating system. Fathom provides a way to look at what and where a requesting machine is, and which URLs those in a given country, or using a particular OS, asked for. It’s pretty neat.

I’m not really sure what one does with these counts. Do you make more pages similar to your most requested page? That’s not a whole lot of fun. You’re letting your audience lead you.

The main value that Fathom adds over just looking at the access_log file is that they interpret things for you. They geolocate IP addresses, and untangle “User Agent” strings. They also seem to eliminate a lot of spiders and crawlers, probably because the spiders and crawlers don’t source https://cdn.usefathom.com/script.js, or at least don’t execute it. Or maybe crawlers and spiders do download and execute script.js, and Fathom is really good (but not perfect) at figuring out which downloaders are crawlers and spiders.

Edit 2025-08-12 I had exactly no trouble canceling my Fathom account. There were a few more clicks than strictly necessary, mainly to keep me from arbitrarily deleting everything. Fathom did not bill me for the next month. It seems that Fathom is a decent company, run competently.