If you ask any marketer what keeps them up at night, chances are high that ad fraud is at the top of that list. And can you blame them? Last year digital ad fraud cost advertisers more than $23 billion globally. Now, these same problems are coming to a TV near you.
That’s right, the same digital capabilities that have marketers salivating over CTV--granular targeting, impression-based measurement, to name a couple - are also opening the doors for fraudsters to start siphoning money away from advertisers. And fraudsters have already stolen an estimated $1.4 billion dollars via CTV using a variety of methods that capitalize on the open ecosystem and the absence of javascript.
Just look at ICEBUCKET, a massive connected-TV ad fraud operation that leveraged server-side ad insertion (SSAI) to send 1.9 billion ad requests per day and spoof two million consumers in 30 different countries. Or DiCaprio, where advertisers were tricked into believing that they were buying CTV ads on premium Roku publishers like CBS News, Fox, and PBS, when in fact they were purchasing inventory on a mobile app.
To top it off, a seven-figure ad fraud scheme deemed Monarch misrepresented inventory to trick marketers like Lexus, Geico, and Jaguar into buying passive experiences like a screensaver app, instead of content from the public domain such as episodes of The Three Stooges or The Andy Griffith Show.
And with Upfront spending down 33% this year, this problem only stands to get worse, as marketers will turn to programmatic to allocate an exponentially larger portion of their marketing budget. So, how can we stop this multi-billion dollar problem that is plaguing CTV? Let’s start by identifying how these flagship CTV fraud schemes could have been stopped.
Don’t Make The Same Mistakes Twice
One of the predominant methods of CTV fraud is SSAI, which is used in roughly 40% of OTT ad serving and can be seen in action via ICEBUCKET. SSAI essentially creates a loophole for fraudsters, allowing them to get access to IP addresses and pretend to send traffic from reputable vendors. And with the OTT environment lacking javascript, detection by most traditional third-party verifiers is next to impossible.
The best way to avoid this on CTV is to work closely with reputable SSAI vendors to understand and validate their credentials. Where applicable, have them log all the channel details, user agent details, and IP addresses on the traffic they send. It’s important to observe trends such as high video completion rates, spikes in traffic, number of devices, bid pricing, etc. All of this needs to be investigated more thoroughly, as this is a telltale sign that may indicate some sort of large-scale fraudulent operation.
Another widely used method is "Platform Mismatch," where fraudsters disguise mobile inventory as CTV ads to capitalize on the higher CPMs. This was the backbone of the Monarch operation and resulted in advertisers paying exponentially higher prices for lower value inventory. The best way to avoid this is by leveraging user agents to identify the device ad requests that are coming, which enables the ability to differentiate between a CTV and a mobile device.
Fraudsters are also using app ID spoofing to have low-quality apps pretend to be something more premium to deliver the ads. And with thousands of streaming apps, it’s easy to buy a knockoff BBC app if you don’t identify a catalog of all the CTV apps and app IDs out there, and actively assess if the app ID sent is accurate.
There are also a variety of other methods of fraud like bots that send invalid, non-human traffic to apps. Or abnormal activity symbolized by things like 24-hour recurring ad requests, constantly changing device ids, and high user volumes, which all raise red flags that need to be looked at. And fraudsters are constantly looking for new ways to siphon money from the ad ecosystem, which is why they are even purchasing data on how humans actually interact with apps so they can more closely emulate it.
CTV fraud is already a multi-billion dollar problem, and if the industry takes a backseat, hands-off approach, this could very quickly escalate beyond control. Especially with CTV expected to fill in for a large portion of the Upfront slump, fraudsters are ready and we must be too. And in order to stop fraud, it’s essential not only to identify where we went wrong in the past, but to be proactively on the lookout for any and all suspicious behavior. It’s not going to be easy, but now is the time to take the necessary steps to prevent the CTV fraud problem from turning into another digital fiasco.