What OTT Can Learn From Mobile In-App When It Comes to Preventing Ad Fraud

OTT platforms are creating new avenues for advertisers to succeed that were difficult to imagine as little as a decade ago. 

For the industry to truly thrive, however, it will need to address the challenges that inevitably come with a growing share of audience time and attention. Within the past year, there’s been a wealth of media coverage suggesting that ad fraud has become a serious issue for OTT platforms. Left unaddressed, it has the potential to stifle growth for the channel, with advertisers opting to spend their media budgets in more mature environments.

The good news is that this isn’t a new phenomenon. Fraud follows money, and as the OTT industry matures, it is experiencing the same growing pains that in-app ecosystems had to start confronting nearly a decade ago. Many of those same lessons are equally applicable to the growth of OTT, starting with the ability to reliably identify fraudsters based on quantifiable behavioral patterns.

Recognizing non-human behaviors

For mobile advertising demand sources, fraud prevention is a two-pronged strategy. The first is the creation of rules that prevent known fraud instances from recurring — a rules-based strategy. The second technique is to get proactive about identifying unusual patterns that are unlikely to represent human behavior. This second approach is essential to stopping fraud before it takes root.

The drawback is that proactive fraud prevention can be significantly more challenging for demand sources. It requires the ability to extensively analyze device IDs, IP addresses, network activity, and the various ways users engage with advertisements. Most importantly, it requires an understanding of what type of behavior could reasonably be accomplished by a human. If your behavioral metrics suggest a user has engaged with five hours of advertising content within fifteen seconds, that’s likely related to fraud. In the OTT space, this could appear as a user who watches videos twenty-four hours a day, seven days a week, without taking a single break.

This understanding isn’t just necessary to spot fraudulent behavior, it’s required to prevent platforms from becoming too cautious. If you’re overzealous in flagging statistically unusual behavior, it’s increasingly likely that you will generate multiple false positives. Eliminating all fraud is costly and next to impossible — the better goal is to aim for the “center mass” of ad fraud that mitigates the majority of instances and makes the space safer for advertisers.

Work to unify fractured environments

Fraudsters aren’t just targeting OTT platforms that are monetized with ad revenue — they’re targeting the fractures that have emerged out of discrete solutions evolving in isolation. Since OTT platforms are relatively new, each brand has its own internally designed advertising standards. This creates a variety of weak points that fraudsters can probe for opportunities.

Mobile ad networks have experienced this trend first hand — Android devices were notoriously prone to fraud because apps can be downloaded from any third-party website or storefront. iOS devices, on the other hand, are always integrated with a unified App Store. This allowed Apple to establish universal standards that made it harder for fraudsters to operate.

Fractured environments aren’t caused by any inherent flaws in device hardware or ad delivery methods. Instead, by offering channels for third-parties to define an experience, the easier it is for malicious agents to take advantage. The more disparate your advertising environment is, the less transparent each interaction becomes -- and transparency is essential for reducing fraud in any meaningful way.

Unfortunately, creating transparency at this scale isn’t something that can be accomplished alone.

Pursue shared standards with competitors

It’s always possible to surpass the technical skills of any individual fraudster. To combat fraud at scale, however, the advertising environment needs to be harmonious — which is easier said than done. A unified ecosystem requires each demand source and advertising partner to adopt universal standards, but finding shared definitions is a challenge in itself.

What should count as an impression? What interactions are measured as clicks? Does watching 90% of a video mean it is “completed” for advertising purposes? The answers might seem obvious, but many ad networks have unique definitions for each metric. Facebook in particular has faced notable criticism for considering a completed video view to mean just three seconds. And that’s before defining what activity constitutes fraud itself.

Admittedly, creating shared standards means demand sources must cooperate with their direct competitors. While transparency is healthy for any ad industry, actually initiating the process can leave individual networks feeling vulnerable. It’s crucial to remember that the consequences of failing to address ad fraud are greater than any individual perceived loss of competitive advantage. The more that demand sources are able to reach a consensus when it comes to advertising standards, the easier it will be for advertisers to invest in them. A rising tide lifts all platforms.

Partner with third-party solutions

Once universal standards are established, how do advertising partners guarantee transparency? Historically, the best approach has been to partner with objective third parties who help ensure everyone is on the same page. For mobile ad networks, these are mobile measurement partners (MMPs) who manage the tracking and attribution of downstream behavior to individual ad campaigns and their associated channels.

Before MMPs, many networks custom-built a solution for recording user data. In practice, this slowed down advertising deployments and disrupted the overall user experience. MMPs offer unified sources of data that streamline the collection and analysis of campaign-specific data, allowing for the first steps to be taken towards standardization.

If OTT platforms follow the same trajectory as in-app advertising, these solutions should materialize very soon. When they do, I have one piece of advice: Don’t ignore them. Never assume your SDK is going to beat the competition. Advertisers want all of their data in one spot, and nothing your SDK offers will beat the security and transparency that access provides. Even giants like Facebook, which provides a first-party SDK, has partnered with MMPs for these exact reasons. Once you start hearing about the same measurement partner from multiple advertisers, it’s probably time to consider working with them yourself!

OTT platforms won’t be able to overcome all advertising fraud overnight, but solutions do exist. By following the lessons of mobile ad networks — seeking out non-human behavior, creating shared standards, and working with third-party solutions — providers will be able to better serve advertisers and drive the continued growth of the environment.

Sarah Chafer is the executive VP of global sales at Tapjoy. She is responsible for global performance & brand sales revenue growth and product adoption. She has been with the company since 2009 and saw them through the transition from web-based advertising to mobile in 2010.

Prior to joining Tapjoy, Sarah was a media director at Medley, Inc. a global digital agency responsible for dating brands. There she was responsible for the user acquisition and affiliate sales team. Sarah has previously held senior level roles at IAC, and Yahoo.

Sarah Chafer

Executive VP, Global Ad Sales, Tapjoy