This is the 12th and final installment in a series of interviews with vendors combating the problem of ad fraud. Read previous interviews with comScore, DoubleVerify, Dstillery, Forensiq, Integral Ad Science, Moat, PubChecker, RTB Asia, Telemetry, Videology and White Ops.
Wasted delivery is not necessarily fraud – but all fraud is wasted delivery.
In other words: While the "why" – be it attributable to botnets, unscrupulous publisher, fat fingers or negligence – is most certainly relevant, what really matters is whether or not an impression hits home.
Alex White, VP of product strategy at ad tech company Sizmek, likes to put everything in a single bucket.
“It’s not a healthy conversation to talk about fraud separately from something like poor placement,” White said. “From an advertiser’s perspective, they want the ad they bought to be shown to a prospect. If that ad shows up below the fold, if it’s a video that has no sound or if it's stuffed into a one-by-one pixel – it’s all just wasted delivery.”
And wasted delivery will always exist – as will fraud, White said. That’s not a pessimistic viewpoint, just a realistic one. Fraud, he said, is an area that needs a dose of realism.
“As long as money can be extracted by perpetrating fraud, someone is going to figure out how to do that,” White said. “That’s why it’s simply a matter of educating the constituents within the ecosystem as to what fraud is and how best to deal with whatever flare-up is flaring up at the time.”
White sat down with AdExchanger.
ALEX WHITE: At its core, Sizmek is an independent ad management platform. We’re not a point solution. We focus on enabling advertisers to deliver their ads and their messages, and, from a reporting standpoint, being able to give them tools to analyze, verify, attribute and collect data.
How does fraud fighting fit in there?
There is a whole industry that’s built up around fraud, verification and attribution, all as separate types of point solutions, when in fact third-party ad serving is exactly that. It was the original verification solution for advertisers to have an independent mechanism for incrementing cost from the publisher’s tool set.
Since rooting out fraud isn’t Sizmek’s main deal, how do you approach the issue?
We don’t jump into things unless they fit squarely in our wheelhouse, and there are two types of fraud that do. One is ghost sites, which is the practice of creating a page that has the look and feel of real content, but that’s only visited by bot traffic. The revenue comes out of the exchange environment. From a pre-bid perspective, we have our Peer39 technology, which is a page-based solution that can tell the difference between the characteristics of a ghost site page and an actual page being visited by a human.
Also on a pre-bid basis, we have the ability to identify spoofing sites, which are sites that come into the exchange as one thing, but then resolve into something totally different. Because we go through the whole chain, we can see exactly what happens.
Finally, Peer39 works in combination with the ad-serving aspects of Sizmek to handle blocking, verification and viewability. We can see a whole variety of attributes, including the language on the page, geography, brand safety, if a video player is set to autoplay, audibility and where a video started and stopped playing.
If an advertiser is using Sizmek, does it also need a point solution?
What’s nice about the point solutions is that they drive innovation throughout the industry. With every advancement Integral Ad Science makes, DoubleVerify is going to follow, etc. We’re all trying to make sure that we’re best-of-breed, which leads to the creation of some really great tools.
But if I were an advertiser, I would take a layered approach. Everyone has different capabilities. A lot of companies out there are trying to be that one-stop solution. There are things that are core to our business and things we’ll need to partner with others on. We have a lot of tools, but we don’t have everything. Do advertisers need multiple solutions? I would say yes and I think everyone who’s honest should also say yes. I would encourage any advertiser to take a deep look at all of the solutions that exist out there and pick the ones that help them achieve the goals they’re looking to achieve.
You also need to be aware of who you’re buying from.
When it comes to inventory, is the onus on publishers or is it more of a buyer beware situation?
This might not be a popular opinion, but I think it’s the advertiser’s responsibility to mitigate their own risk because some publishers are the bad actors in this situation – not all, but certainly some. And there are plenty of tools out there that advertisers can use for protection.
That said, both parties need to work together to create an environment that’s as fraud-free as possible.
How does that tie into your stance on wasted delivery?
Wasted delivery is bigger than just fraud. Fraud is one aspect of wasted delivery, but it’s not the end all, be all. That’s why advertisers need to have a good understanding of what they’re buying when they’re buying media. If you don’t know what you’re getting, then you have a big problem.
If you want to put the blinders on and just buy cheap inventory and not worry about it, that’s fine, but then don’t be upset when your ad shows up, or you’re told that your ad had shown up, in an environment that’s either fraudulent or porn or some other nonaligned environment for the brand message.
Who’s at fault when it comes to waste?
It’s not necessarily anyone’s fault. I see it as a symptom of the industry that has matured around us. Wasted delivery can take many different forms. If a video autoplays below the fold with no sound, that’s perhaps not fraud, but it’s definitely not a good user experience or a good situation for the advertiser.
Publishers are pushing the boundaries of what they can do in order to get paid. I wouldn’t think any of the top 250 publishers are doing anything like this, but once you get beyond that list, there are tens of thousands of publishers out there, and many of them might. It’s very easy to buy inventory that’s going to end up as wasted delivery.
Can you share an example?
At one point, one of our major CPG clients found that over a period of about five weeks nearly half of the the more than 1 billion impressions served during a video campaign of theirs was considered to be wasted inventory. A quarter of the impressions had no video start, 123 million impressions were not brand safe, more than 146 million impressions weren’t even in a video player and more than 17 million impressions were below the fold.
There was also a subset that were in the wrong geo, about 10 million impressions. That’s a very small slice of the pie, but for this particular advertiser, it represented $55,000. That’s a fairly sizeable amount for a relatively small slice.
Nearly 50% of the inventory was what we would call wasted – not fraud, but wasted. The definition of fraud is pretty cut and dried, but advertisers have different opinions about what represents wasted delivery or brand safety to them.
So fraud is clear, but wasted delivery can be nuanced, depending on the advertiser?
I came from the Peer39 side and our belief was always to give people the building blocks to assemble the profile that makes sense for them. Items like “religion” and “politics” show up as brand safety categories because some brands have sensitivity to those topics. But that’s just contextual. For example, some advertisers are fine with gambling content and others aren’t. It’s all subjective.
Is the media engaging in yellow journalism on the ad fraud issue?
Every reporter I speak to on the subject doesn’t disagree with the assessment that this is happening. The media jumps on numbers that are being thrown out and those numbers are typically being produced by the providers of the fraud solutions themselves in order to scare advertisers.
But part of the problem is that there are no real numbers out there. You can’t yet put a number on fraud.
What about what the coalition study between the ANA and White Ops? They’re working together to try and finally determine that number. Results are expected soon.
What they’re trying to do is awesome and it’ll give an indication, but I don’t think it will be representative of the entire industry. That number will be the level of fraud among the advertisers they had the pleasure of working with. Granted, they’re some of the largest advertisers in the world, but they don’t account for 100% of the environments that others are operating in. That said, I do applaud them for getting multiple steps closer to a real answer as opposed to random companies throwing out big numbers to get quoted in news articles.
Which ties right back into the fraud feedback loop.
I said earlier that it's up to the advertiser to stay informed, but how can the advertiser do that when there's misinformation, bad stats and fear mongering flying around everywhere.