At Spotify, we believe a view should be sight and sound and motion, on screen, through to the end.
If a tree falls in a forest and nobody is around to hear it, does it make a sound?
Likewise, if a digital video ad delivers and is not viewable or audible at the end, is it a completed view?
This begs the ultimate question: what is a media buyer really receiving when purchasing guaranteed completions?
Digital advertising’s CPCV (Cost Per Completed View) pricing model was first introduced in the late 2000s in order for media buyers to reconcile poor industry completion rates that have since remained steady in the mid-60% range worldwide.
Since a “video view” was never defined further than simply an ad “start” (like any other digital impression) there was no way to guarantee whether or not your message ran its course. Enter CPCV, which was heavily promoted by video ad networks as a way to solve the problem.
Note: my intention here is not to debate arbitrage models or the higher effective cost of CPCV. What I think deserves scrutiny is how we define that completion. Buyers deserve to get what they pay for.
Completions should be qualified with whether the message was actually visible and audible at the point of its conclusion.
Moat has a metric for this called AVOC: Audible, Viewable On Completion. Add an “H” to this acronym—representing the confirmation the ad was delivered to a Human—and you have: HAVOC.
The Oracle-acquired company just released their Q3 benchmarks last week. The numbers illuminate the poor state of viewability across the board…but nothing is more frightening than the low AVOC numbers.
Globally, Moat reports that on desktop browsers, video ads get viewed by real people only 62.2% of the time. (Based on the MRC standard for views: 2 seconds with 50% pixels in view.) This number drops to 49.2% on mobile in-app platforms.
Now, we can debate the standard all we want, but even at this low bar, over two-thirds of all video impressions fall short. No wonder eMarketer reports that viewability is the #1 KPI for digital media buyers.
Looking at AVOC, the story gets much, much more concerning. In Q3, Moat’s reported global benchmark was 31.1% for desktop and 24.7% for mobile in-app. This means that over 75% of the time, a person isn’t listening to and viewing your message in the final seconds of the spot.
So, if an end ping occurs and a publisher counts a completed view while the sound is off or another app is active, can they really call it a completed “view?”
At Spotify, we believe a view should be sight and sound and motion, on screen, through to the end. Moat’s Human + AVOC Metric (HAVOC) is the best proxy to quantify whether someone used their eyes and ears to consume your message all the way through.
That’s why earlier this year, we began offering a guarantee that completions bought on CPCV would be HAVOC-verified by Moat. Advertisers only pay for views that Moat could measure and certify were audible and viewable during the final two seconds of the creative.
The market responded so positively that we’ve now extended this offer to our new Mobile Video Takeover unit. So, if a user locks their phone, receives a call, or minimizes the app during the middle of an ad, that view won’t be certified as HAVOC and will not be billable.
In today’s environment, transparency is critical to establishing and maintaining trust between media buyers and sellers. That’s why we work with the leading third-party verification partners to prove our clients are receiving media that’s optimized to drive the best results.
The MRC standard for video viewability, like it or not, is still the primary metric advertisers use to measure media success. However, by applying it in a slightly differently way, we can better define what a completed view actually is and how we can all transact on it.
Brian Danzis is the Head of Global Video Monetization at Spotify.