Yes, Viewability Measurement in FouAnalytics

Yes, we measure viewability in FouAnalytics. And we have security measures in place to prevent tampering, like what was documented with Newsweek in 2018 — Newsweek Media Group Websites Ran Malicious Code to Make All Ads Appear to be Viewable.

Viewability Standard - Definition

The industry definition of a viewable ad is “50% of the pixels in view for 1 second.” Why isn’t it 60% or 100%? Why isn’t it half a second or 1.5 seconds? Those don’t really matter anyway, because the current standard was just some arbitrary threshold the industry decided to use after 7 years of committee discussions. And it is utterly useless to advertisers anyway, because fake sites always have 100% viewability (due to cheating) and real sites have lower average viewability (because it’s reality). A page that has 4 ads on it, with 3 above the fold and 1 below the fold, will have a 75% average viewability. That is not a bad thing, especially if the page is on a real publisher’s site. When media agencies stupidly sent out press releases saying they would only by 100% viewable ads, they didn’t realize that incentivized the fraudsters to use malicious code to make all their ads viewable; and it simultaneously disadvantaged real publishers because in an instant, 25% of their ad inventory became unsellable.

Viewability in Practice, and in FouAnalytics

Of course, it makes sense to want ads to be viewable (opportunity to be seen) but optimizing for 100% viewable is not wise, and in fact will lead to higher risk of buying ads on fraudulent sites. So in practice you should have basic checks to make sure your viewability is not absurdly low, but it doesn’t have to be 100%, and never will on real publishers’ sites anyway due to basic page layouts.

Viewability in FouAnalytics is measured using three simple things: 1) does the ad slot intersect with the viewport (visible portion of the browser inside the navigation bar and scroll bars), 2) is the browser minimized and browser tab active, and 3) is the browser the active application? FouAnalytics marks an ad viewable if 50% or more of the ad slot intersects with the viewport, the browser is not minimized and the tab is the active one, and finally the browser is not behind some other application like Powerpoint, etc. Note that we deliberately ignore the 1 second because if the ad doesn’t start out as visible, it is unlikely it will become viewable (user scrolls down the page to see it) before the user leaves the page anyway.

You can find this viewable parameter in the data grids in your FouAnalytics dashboard for in-ad measurement. Network-wide we are seeing about 22% overall average viewability for display ads in North America. Note that when we account for the 1 second in the “viewable-MRC” data grid, it is a bit less at 18%. This makes sense since the ad impressions that also meet the 1 second rule is a subset of the viewable=1 impressions.

One key point of advice. Don’t use viewability metrics in the absolute. That means the exact number 21.6% is not what is most important. It is RELATIVE viewability between channels, supply sources, or publishers that is important. For example, if one supply source has a higher average (untampered) viewability, then allocate more budget to that one. Avoid sources that have 100% viewability 100% of the time, as those are likely faked (let me know so I can check how they are doing that).

If you’re still paying for viewability measurements from other ad tech vendors, you’re wasting your money. Check when those contracts expire so you can stop wasting money.