I am using Plausible, a month with them and enjoying it a lot.
Plausible was covered in detail in the last week’s article. That’s how I found out about it and am now happily using it for kitspace.org.
Glad to hear that Kaspar, thanks! That was such a nice article that introduced couple of great GA alternatives to many new people, really appreciative of Ben’s coverage!
Thanks very much for these nice words Guillermo!
(I’m the co-founder of Plausible)
The number one thing on this list ought to be “ask yourself if you really need analytics, and what for”. If you’re an individual running a website, analytics likely serve no functional purpose besides releasing a little dopamine in your head when numbers go up.
If you like simplicity and open source software, I like GoatCounter
I’m still at a loss for a good distributed log analyser.
GoAccess has most of what I/we want, but it’s got no support for e.g. storing the accumulated metrics in e.g. Redis, or SQL, or what have you, and thus it’s not really fantastic for anything with > 1 web server.
So far I’m still tossing up if I should just write one myself.
Why not try submitting a patch upstream? I’m sure other GoAccess users have similar needs.
GoAccess is written in C, which is most definitely not in my wheel house. Ok, so, why not just learn some more?
The previous time I took the effort to make a patch against a C codebase (not that being C is the specific thing, but being a significant effort required due to my inexperience, compared to either a language I am familiar with, or a person familiar with C) was also, “scratching my own itch” - I submitted the PR 1 year and 19 days ago. So far, the ‘owner’ of the project hasn’t responded to the PR at all.
Previous PRs haven’t all necessarily been completely ignored like that, but I’ve had more multi-month+ delays in getting any kind of feedback/activity, than I have “positive” engagements.
I guess what I’m saying is: the effort involved for me is going to be quite high, and having to maintain a fork of a tool written in C (if there’s zero upstream action/interest) isn’t really on my “hey this sounds like it’d be fun” list.
Edit: and I should add - I’m not “negative” on Open Source. The stuff I write for my company is specifically all OSS, and I have obviously had some reasonably good experiences with upstream projects. But the often bandied “if projects aren’t on GitHub they wont get contributors because it’s harder to find/no network effect” line, to me misses a significant aspect. This isn’t PR field of dreams, where “if you PR, they will merge it”. Just because someone submits a PR doesn’t mean there’s any specific likelihood it’ll ever be looked at.
I’ve had similar experiences. One of my own PRs took a year and a half to get accepted, and I was just updating documentation!
I didn’t mean to suggest that you should rush into writing code. Open an issue outlining the problem first to gauge interest. Make sure you note your willingness to contribute code, but don’t write a line until one of the maintainers makes it clear that they’re on board. Of course, if you really don’t want to take the time, then don’t! It’s your life; live it the way you want to!
Do any of these services work well with multi-tenancy? Let’s say I have a SaaS app that puts clients on subdomains. I’d like to access analytics in isolation (per-client) but also in aggregate. I think I can achieve that with GA with a lot of setup.
Not mentioned here are https://metrical.xyz/, https://usefathom.com/, and http://simpleanalytics.io/. All are privacy-focused.
We briefly mentioned Simple Analytics and Fathom in last week’s article, Lightweight alternatives to Google Analytics. However, we didn’t focus on those tools because they’re not open source (Fathom was, but version 2 is not; version 1 is now called Fathom Lite). LWN is big on open source, not so much on proprietary tools.
Equally Offen is another alternative who market themselves on being ‘fair’. Haven’t used them personally however.