Kinja Data Team

it's A/B test time, my dudes

About a year ago, ad-blockers were the existential threat du jour for online publishers. Even relatively sober publishers like The Economist were writing articles about Adblockalypse. Even we got in on the fun.

A year later, concern and interest in the topic is waning:


Recent articles about ad-blocking are about as likely to be about Facebook fighting blockers as they are about an impending threat to publishing. It’s possible that people are just resigned to ad-blocking and have moved on, but I think the industry is treating this as less of an issue than last year.

What makes this change interesting is that more and more people are installing and using ad-blockers. When I recently took a look at our ad-blocking numbers to see how they’ve changed over the last year, I noticed that there was an increase in their rates on all device types.

But despite the increase in ad-blocking on each device, our overall ad-blocking rate was going down.


I was pretty confused about the numbers, but after a little investigation, it became clear what’s happening: the shift to mobile traffic is dominating the increase in ad-blocker usage.


The move to mobile traffic*, at least for our network, has picked up over the last year: we’ve gone from about an even number of mobile and desktop sessions to a 3:2 ratio:


Since ad-blocking rates on mobile are still just a fraction of what we see on desktop, we see our overall ad-block rates falling quickly, even as adoption on each device platform increases.

There is a technical name for this phenomenon of trends reversing when groups of records are combined: Simpson’s Paradox. Statisticians and data analysts generally love talking about unusual, counter-intuitive findings, and this is one of the best known. I’m pretty sure it’s the first time I’ve ever seen it so clearly in my career though.


Publishers are probably right to focus on other issues at the moment: concerns about relationships with platforms and the move to programmatic advertising are probably more crucial to their day-to-day. Eventually, when mobile ad-blocking rates start moving up in the US and Europe (like they have in Asia) I’m sure we’ll be hearing about this more.

If you want to learn more about Simpson’s Paradox, including its famous use in the 1973 UC Berkeley discrimination case (and why it’s neither Simpson’s nor a paradox), I strongly recommend this fun visualization.


*The move to mobile also offers some challenges in tracking ad-blocking usage: we think mobile ad-blockers are more likely to block our tracking than desktop ones. Also, restrictions on certain trackers on Facebook Instant Article and Google’s Accelerated Mobile Pages is pushing these numbers lower. However, even taking these data deficiencies into account, the overall trend I’m discussing holds up.

Share This Story

Get our newsletter