JOHN MONTGOMERY: When I got this job in September, my first question was how to get a consistent viewability standard. If we have strict viewability standards, would we get enough inventory in each [international] market?
Our standards in video are identical globally as in the US. But rather than pay against viewable impressions internationally, we’ll optimize against them. In the US, we have agreements with 150-odd partners, who supply us with viewable inventory that meets our standards. We pay based on viewable impressions.
But in Poland, probably only 10% to 15% of video impressions are viewable. So when we go into that market, we say: "These are our standards. We know you’re not there yet, but we’re going to optimize against those standards. So if you want more money from our clients, give us more viewable inventory."
In a year and a half, we’ll have more than enough inventory to reconcile against views.
When GroupM announced its viewability standard about two years ago, there was backlash. Did you ever have to adjust to accommodate some of the ill will that standard engendered?
The original standard was a good basic benchmark, but if clients are to move money from television into online video, they need [the GroupM standard]. It was controversial because no one expected us to do that, and Randy Rothenberg is still irritated with me.
But I think GroupM’s efforts have lifted the overall standard of viewability. So if there was ill will, it was worth it though it was never intended. It was worth it because more than 60% of inventory out there is viewable by GroupM standards and more than 85% is viewable against MRC standards. When we started, only 50% was viewable by MRC standards, and only 20% was viewable by GroupM standards.
And premium publishers have benefitted the most.
IPG and Integral Ad Science have been talking about how even if an ad isn’t fully viewable, it still has value. What’s your take on that?
Boz [Andrew Bosworth] from Facebook was asked his definition of a viewable ad. He said anything with more than zero pixels and seen for more than zero seconds has value.
Except it has no value to us, or to advertisers. If 50% of an ad is viewed, which 50% is better? The part with the logo? The part without the logo?
If I’m moving money into digital, I would want to know that at least all my ad was in view. And if we can negotiate with vendors to deliver that, why would we not?
I had an argument with a person the other day, who said video with the sound off has value.
Was that a Facebook person?
It wasn’t a Facebook person. But they said you could make ads with subtitles. But that’s a compromise. Audio-visual needs to be together. If we can negotiate that on behalf of our clients, and if vendors agree to it, why on earth would we not do it?
So do you pay a different rate for a sound off ad?
We might. We can’t not use Facebook, though some clients have said they won’t use it if it’s autoplay with sound off. Some will use it for display, but not video.
But some compromise. They’ll make sure [something happens] in the first five seconds, so it attracts your attention and you turn the sound on. If we have to create gimmicks and tricks to get things viewed, I won’t say it has no value, but it has less value in getting a basic ad seen and acknowledged.
Isn’t there value though in understanding the nuances? For instance, maybe think of muted video as a rich media ad instead of video. It’s more valuable than display but less valuable than video.
Except it’s not being priced that way. But your contention is correct: If you judge an ad based on its value, say you pay 10% of the value of an ad that’s user-initiated with the sound on and that’s viewed entirely, then that’s the right discussion to have. Especially as we go into new video.
We’re announcing, in the next couple of weeks, our standards for mobile and new video.
What’s “new video?”
That’s news feed video, Snapchat, Twitter, even Facebook Mobile, where our premium video standards don’t apply. People are scrolling through it, but they’re doing so very quickly. If it activates when it’s been scrolled through, is that user initiated or not?
At some point in time, that ad will be 100% in view for a microsecond. But does that give it opportunity to be seen? We need to figure out how long the ad needs to be in frame for someone to register what kind of video it was.
So we’ll have standards that adapt to that. We’re figuring that out now.
How would viewablity and brand safety work in things like addressable TV, or Alexa – where by definition nothing is in view?
We’re addressing those new issues as they come along. But frankly, there’s very low volume at the moment, which makes them academic discussions.
We didn’t tackle new video until it became necessary. But now we are because there’s real volume.
Facebook and Google are the ones who are pushing to get that TV budget. How receptive have they been to GroupM’s viewability initiatives?
Google has come around with TrueView. It’s high priced – you pay more for it – and Google has been more open to measurement recently.
Facebook is slow, but they allow Moat tags and IAS tags. And they’re making an enormous amount of money and are therefore in a position of strength right now.
But we have big clients who aren’t spending any money on Facebook at all. We’re having discussions with Facebook all the time and we think they’re coming around slowly, especially given some of the recent malfunctions they’ve had [with measurement].
Have those measurement issues given GroupM leverage? Or is Facebook so big that those hiccups don’t matter?
I think from a financial perspective, they’ve hardly noticed. But from an image perspective, it’s got their attention. It’s not about leverage, it’s about what’s the right thing to do, and those issues have shown them what’s the right thing to do.
Is there a point when the image issue becomes a financial issue?
That’s ultimately the threat for them, but they haven’t gotten there yet. But if they don’t do anything about it, it’s certainly a possibility. I know their recalcitrance on viewability standards have cost them money from our larger clients. There’s no doubt about it.
But we don’t want to talk about leverage. It’s an amazing medium with an amazing approach. Still, we worry about the walled garden. We want viewability and to measure more accurately.
Isn’t there an extent to which the walled garden approach benefits GroupM? With your buying capabilities, you can get through all of those gardens, which gives you more power.
No. To some degree you might be right, but we don’t see into those gardens the way we need to.