What should advertisers be on the lookout for?
Marketers want to know their impressions are human and viewable, and not coming from the variety of ways that fraud can be perpetrated. Really what we’re talking about these days are the hygienics of streaming. It’s mission critical to the survival of streaming and the survival of television.
There are several types of CTV fraud to be aware of. Geographic misrepresentation happens when an unscrupulous arbitrager may sell ads they claim are being shown to users in the U.S., but instead will show the ads in a developing country to take advantage of lower CPMs, as an example. With device-based fraud, there’s a process known as “looping,” where a single connected device will show an implausibly high number of ads during a given time period — for example, 60 ads in a single minute. App-based fraud occurs when an ad-supported OTT app shows a consistently high rate of activity around the clock. A less sophisticated kind of fraud that can frustrate marketers is seller-based deception—sellers selling impressions that are represented as one thing and end up being quite another. With some of the reporting clients can get, there’s often a bucket that says ‘other.’ If you’re a client, you want to ask what’s in that bucket, where do they come from?
Sellers also need to be educated as sometimes they are engaging in deception without even knowing it. There’s truly an incumbency on the industry to make sure our sellers know what they’re selling. A lot of vendors have entered the OTT and CTV sales game, and it’s not as simple as, ‘Hey buy my impressions.’ While it could be a seemingly innocuous representation, it’s still something to look out for.
Are there other losses that are incurred when ad fraud takes place?
There’s of course the monetary loss, but there’s also the opportunity cost loss. For instance, a marketer needs to know that the impressions they bought were delivered as ordered. An advertiser’s campaign is carefully planned to do something very specific at a specific point of time. Fraud not only makes it net effectively more expensive, but also robs you of the actual opportunity cost of that campaign in time.
How is Premion’s approach different than that of other OTT advertising platforms?
We employ several layers of defense that leverage forward-thinking technologies. For several years we have been working with MadHive. They are an ad tech partner for us, and they’re recognized leaders in addressing CTV fraud. They are using machine learning in this area and automated content recognition, and they’re also using cryptography not for “code breaking”, but to look for irregular patterns in traffic and inventory verification
In addition to working with MadHive, Premion’s business model keeps us at the forefront of ad fraud protection. Anything we do is a direct deal with broadcast and publisher partners. We are all about direct relationships; we are not arbitraging the open exchange. We are really focused on leaving no room for any suspicious, obfuscated, fraudulent activity. We also have no user-generated content and we put a premium on understanding and knowing our source inventory.