Fraud in Mobile Video


When it comes to mobile video, do you really know if the ads you are serving are actually hitting your intended audience? This question comes to mind as there has been a lot of talk recently about mobile video fraud in the ad tech industry. While there are already multiple ways in place to track fraud for display, social media, search, etc., mobile has been a completely different beast for tracking.

The conversation around mobile fraud really got kicked into high gear on August 28th, 2015 when Rob Leathern, former CEO of Optimal Inc, wrote a post titled “The Mobile Video Ad Lie”, where he threw some harsh accusations at the mobile video industry. Without going into too much detail, he noticed very long load times for publisher mobile sites when there were only a few static ads, and so he decided to run his iPhone’s traffic through a Charles Proxy to register the calls that were being made on his mobile phone. A test of the mobile site revealed that there were 900 HTTP/HTTPS calls in a five minute time frame for about 10.3 megabytes worth of data. There were also a total of 291 pixel sync calls, which are the sharing of his IP address across different publishers and data providers. If you are interested in reading his full analysis, check it out here.

One of the main culprits called out in the article and a follow up piece written by Mike Nolet, former CTO of AppNexus, was PLYMedia. PLYMedia was sending out multiple ad calls to Facebook’s LiveRail and AOL’s Adaptv 2. The conclusion was that mobile video ads, being premium inventory, are basically being “served” behind standard mobile banners to increase mobile ad revenue. This is costing advertisers a lot of money, and is yet another blemish on the ad tech industry’s record. PLYMedia’s CEO responded by saying it was only a test of a new service that is currently in beta 3 and that they are working on fixing the issue, but that the different calls were not anything nefarious. Regardless, Rob Leathern’s test revealed that there was a very large amount of data being transferred when there shouldn’t have been.

Whether or not you believe in Rob and Mike’s tests, there needs to be a lot more done in terms of preventing mobile fraud. Just Media’s internal DSP partner MediaMath was also implicated in the mobile fraud and provided their own comments about the articles by Rob and Mike4. They are doing their best to fight fraud, which is important as ultimately our success lies in being able to trust the exchanges we have access to through our DSPs. MediaMath has done a good job of making sure inventory is indeed what we believe we are paying for and is swift to respond to any suspicious practices.

In the end, the best way to combat mobile video fraud is to hold exchanges, DSPs, and publishers accountable by using data. For this reason we recommend to all of our clients that they start moving away from measuring campaign success solely on CPMs, CPCs, impressions and clicks. Instead, we need to focus on the backend metrics: users landing on the client’s pages, engagement with other pages after landing, and ultimately conversions. When we shift away from the frontend metrics we will be able to better identify those that are fraudulent, as ultimately we will stop running campaigns with those that only worry about impressions, clicks and views.

Sebastien Damas
Just Media, Inc.



Share This Article