I’ve been using Google Analytics on this website since its inception, and according to W3Techs, 56.4 percent of sites on the internet do too. I use the data collected from Google Analytics to see which posts are popular, and features are used most often. If a particular article is more popular than others, I’ll make sure to write more about that subject in the future. If a specific feature is rarely used, I’ll most likely remove it.
My confidence in client-side analytics all changed the day I enabled Netlify’s server-side analytics on my website. After 20 or so minutes, I started seeing the data flow in, and I was utterly blown away with the results. On average, I was seeing an average of 600 page views and 400 unique visitors a month in Google analytics. Not much, but something. On the other hand, Netlify Analytics showed that I had 4,152 pageviews and 1,558 unique visitors in the past month!
What’s going on here? To me, this seems pretty evident that adblockers are pretty damn popular, and most of my traffic clearly uses one. I’m guilty of this too since I use an ad blocker system-wide on both my computer and phone. For those that are interested, my computer blocks ads by modifying the systems hosts file, and my phone uses Blokada. Blocking system-wide is the only way to prevent tracking through all applications.
Nonetheless, never did I expect so much of a gap in the data. This revelation made me sit down and think about all the metrics based decisions I’ve made in the past. Even at work, every single interaction is accounted for, or at least I thought the majority of them were. Features and roadmaps are built around knowing this type of information. Is it possible we are all making the wrong thing because the data is so skewed?
Almost every browser except Google Chrome these days promote themselves as an ad-free, tracker-free browser. Chrome has a proposal to deprecate the webRequest API, which directly impacts how ad blockers operate. I firmly believe that once peoples ad blockers stop working, they will flee to an alternative like Firefox, Brave, Safari, etc. No, not all people will switch, but I think a chunk of their userbase will. And once Chrome begins to lose market share, the analytic results will be even more misrepresentative.
I’ve been ranting about how bad the data is, but you’re probably wondering what we can do about it? I believe Netlify is heading in the right direction by collecting the data server-side. WIth only server-side analytics, you lose out on certain events you would typically track client-side, but in turn, you will have much more accurate data. This might be more beneficial to your business, or you can settle for inaccurate results. In my case, page views and being able to determine which are the most popular is good enough for me.
Perhaps you don’t have a static website as I do, and you have an app powered by a REST/GraphQL/SOAP/or whatever API. You can track each request that comes in, extract the desired information, and then send it to your analytics provider. Processing all data confined within the bounds of your server allows you to bypass any adblocker.
Since my discovery, I have since removed Google Analytics from this website. I believe I can live without the finer details and settle for the limited data collected from Netlify. On the upside, I promote and respect your privacy since there are no longer trackers of any kind. I believe there needs to be a trade-off between invasively tracking everything about a user, and the bare minimum metrics to make a product better. Imagine if every single person and action was untracked. The internet would be a much different place than it is now.