Can Social Help Solve Video Advertising Ad Fraud?
By measuring the social activity taking place around a video and using that data in combination with raw viewing statistics, marketers can gain a better picture of how their video ad is performing.
Viewability has been a hot topic over the last few months. From the early investigations of AdWeek‘s Mike Shields (now at the WSJ) to this most recent New York Times article to the latest rumblings at OMMA RTB, viewability has been top of mind for publishers and advertisers alike. And while plenty of much-needed activity remains directed at reducing fraud and fighting bots, it’s also necessary to now expand the scope to video engagement.
Simply put: Pure viewing metrics don’t tell the whole story. We need not only verification of these views, but also an indication of their effectiveness. For that we need to look outside traditional metrics to provide advertisers the engagement stats they need to make their ad buying decisions.
To that end, established social stats can add much-needed color and context to the raw viewing stats we’re limited to today. I’m referring to stats such as commenting on the video, ranking the video, sharing to Facebook and Twitter, and so on. These are stats that refer to the content which video ads are placed against, and as such obviously only works for pre-roll video ads. But since pre-roll makes up a vast majority of video advertising buying today, it seems like a valid place to start.
Measuring social activity taking place around a video and in addition to the raw viewing stats tells us far more than what viewability ever could. Social activity gives us data such as user intent, engagement, and enjoyment. Seven years ago, we began ranking videos by social data rather than view count, and subsequently updated our search algorithms to weigh the same more highly in our results. The response from viewers has been overwhelmingly positive, as it’s helped them more easily find the most entertaining content.
Advertisers deserve access to the same data and analytics. Monitoring social activity is a good detector for fraud. For example, one indicator that we use with our TruTraffic service is the view-to-share ratio – if that is out of whack, it’s likely a bot racking up views rather than humans.
Social data around videos is generated in great volume. Facebook posts that include video receive 100 percent more engagement than average posts, Twitter drives more than 700 YouTube videos every minute, more than half of 25- to 54-year-olds share videos online, mobile video ads with social media links drive 36 percent higher engagement, and 100 million YouTube viewers take some kind of social action (like, share, comment) every week.
The data is out there. Let’s use it.
