Read on Twitter
Our client-side integration with Mixpanel lost ~50% of events to ad blockers. I know this because of our "Upgrade To Paid" event: the actual subscription count in Stripe was always around half of what showed up in Mixpanel. Now I'm removing all client-side analytics.
Google Analytics also misses at least 50% of traffic. I installed Fathom Analytics (a privacy-friendly web analytics product) a few days ago. It shows about twice as many pageviews as GA did. Probably the same 50% of users who are running ad blockers.
So what's happening with all of the products that use client-side Mixpanel and GA? I was using both in their standard, default configurations. Most of these products' users just have wildly incorrect numbers? We probably have much higher ad blocker use rates than most, but still!
Even if you only lose 20% of potential analytics to ad blockers, it's not a random 20%! It's a well-defined demographic: technically savvy users who install ad blockers. That group is invisible to your analytics. And you steer a company based on that? This all seems very bad.
Mistake in the first tweet in this thread: subscriber count in Stripe was always roughly double what Mixpanel showed, not half.
@fathomanalytics has been great so far. It only does the simple stuff, so you probably need something like Mixpanel too. But Fathom is so simple and clear. Their DNS setup process was especially great. Comparing this to GA... Google should be embarrassed of themselves.
Oops... I meant @usefathom.
