CTV Ad Fraud—What Advertisers Need to Know in 2023
The economy continues to tiptoe around an impending recession, which means advertisers are even more focused on making every dollar count. This is especially true in premium, high CPM environments like the CTV space, where ad spend exceeded $19 billion in 2022.
But to increase ad spend efficiency, the industry must address the neverending fraud and viewability problems that continue to run rampant.
What does CTV ad fraud look like these days?
And how are adtech platforms and publishers battling it?
There are some easy answers…and some ongoing questions to dig into this year.
The state of CTV ad fraud and viewability issues
Issues of ad fraud and viewability have always been a challenge for connected TV (CTV) advertising, but they inevitably become hot topics when there’s economic uncertainty — and rightly so.
The jury is still out on what digital ad fraud cost the global industry in 2022, but estimates range from $60B to $120B (either way, that’s a LOT). There’s little data on how much of this fraud comes from CTV advertising. But thanks to more limited quality inventory in the CTV market, CTV ads have a higher CPM. This creates a bigger opportunity for fraudsters looking to make a buck — which means we’re likely to see increasing fraud schemes looking to steal these ad dollars.
And advertisers aren’t the only victims.
"CTV ad fraud doesn’t just impact advertiser ad spend, it also impacts publishers’ bottom lines,” said Mimi Wotring, Senior Vice President of Publisher Sales and Client Services at DoubleVerify. “Schemes like MultiTerra and SneakyTerra siphon millions of dollars of revenue from high-quality CTV publishers without their knowledge.”
According to DoubleVerify, ad fraud could be costing publishers close to $150M annually.
The heat on the topic turned up further last year with the report on CTV ads that continued to run after the TV was turned off. And in 2023, we can expect to see new and more sophisticated forms of CTV fraud emerge, as bad actors adapt to changing technologies and ad formats. For example, there may be an increase in fraud related to programmatic advertising on CTV platforms, including techniques like bid caching and bid duplication.
“CTV represents a fantastic opportunity to deliver big-screen, targeted advertising to audiences that have switched away from linear delivery,” explains Anything is Possible CEO Sam Fenton-Elstone in an article on The Drum. “But if we don’t control fraud now, it will become as murky as the display ecosystem is already.”
Recession or not, the industry needs a lasting solution to CTV ad fraud.
Adtech platforms are upping their ad fraud game
In response to this need, DoubleVerify recently released a new viewability offering that allows advertisers to verify that a CTV ad was seen by a real person in a brand-safe environment. Clients of DV and their partners (like Madhive) can leverage their segments within their tools.
But this solution only reports out on fraud and wasted ad spend after the fact. Stakeholders across the supply chain are calling for a more holistic decisioning engine to weed out the fraud in real time.
In other words, we need a solution to avoid fraud in the first place.
But weeding out fraud in real-time isn’t easy — it takes a massive amount of computer science and data processing capabilities.
At Madhive, our bidder makes ad decisions on as many as 10 million requests per second, each with a 20-millisecond response time. Every millisecond counts and could provide meaningful, incremental value for businesses. That’s why it’s crucial we enable fraud capabilities without sacrificing efficiency.
This is a challenge that we’re actively working to solve — and we’re uniquely equipped to solve it. The Madhive infrastructure is purpose-built for streaming, which allows us to process massive data sets in real time. Because we started in CTV, the Madhive platform has exponential amounts of historical data and knows what to look for to identify a valid request.
Madhive has also seen incredible growth over the past few years, allowing us to pressure test and scale our data computing gradually. Now, we’re powering billions of local and regional streaming ad placements for clients like Fox, Scripps, TEGNA, and thousands of their advertisers every week.
Implementing an ad fraud “toolbelt”
While many of the tools designed to fight ad fraud are point solutions that address single issues, advertisers and publishers would do better to implement a more holistic “fraud toolbelt”. Carpenters show up to work with a hammer, screwdriver, pliers, and a variety of other tools in their toolbelt. This is the same approach advertisers need to take when it comes to fraud and viewability problems.
Adtech platforms that take this approach consist of several technology layers to combat fraud in multiple ways. For example:
Using direct connections to publishers and built-in mechanisms to detect when the TV is on based on data, such as session duration.
Leveraging tools like ACR data to verify that ads actually hit the screen and panel data to detect behavior signals that show when a person may no longer be present in front of the TV.
Using machine learning and AI to identify premium placements based on historical data analysis.
Platforms that take this toolbelt approach — like Madhive — are detecting the fidelity of impressions throughout the campaign process and making optimizations in real-time. Clients are also able to integrate third-party partners like DoubleVerify to further combat fraud.
Fraudsters will keep trying to disrupt the supply chain, but having a robust toolbelt of solutions that work together and are constantly learning and improving will help weed out the fraud with confidence. When you invest in this sort of approach, you won’t need to rely on point solutions.
Related content:TAG and ad fraud: Why we got TAG certified (again)
Why CTV ad fraud should not deter advertisers
We know there is ad fraud, which is why we have built-in multiple lines of protection for our clients to help them weed out the bad actors. But obsessing over weeding out every single slightly suspicious impression can potentially do more harm than good.
Madhive CEO Adam Helfgott explains, “We’ve seen false positives for ‘suspicious inventory’ result in fraud detection companies eliminating a whole swath of demos, causing tens of millions of impressions on premium supply partners like Roku and A&E.”
This lack of reach and frequency can hurt business objectives even more than fraud.
Instead of becoming fixated on fraud, Helfgott advises that advertisers should stay the course and work backward from baseline attribution:
“Define the basic benchmarks and attribution for your campaigns, and this will inform whether or not your media is working.”
Want to learn more about how Madhive is combatting CTV ad fraud? Check out: How to prevent ad fraud in OTT (5 methods we use)