While bot traffic and viewability are industry-wide problems, only a handful of companies have borne the brunt in the mainstream media. One of those companies was Blinkx, a digital media technology, distribution and monetization platform that focuses mainly on video. In March of this year, Benjamin Edelman, Associate Professor of Business Administration at Harvard Business School, made several allegations about the company, many of which the company has since refuted. Since then, the company has been working on an ambitious attempt to eliminate bot traffic and viewability problems across all of its inventory. Here Donald Hamilton (DH), SVP of Blinkx Media, and Erol Soyer (ES), Managing Director of International at Forensiq, a verification company, discuss what went wrong and what the company has been doing to put things right.
It became public earlier in the year that Blinkx was having problems with fraud and viewability earlier in the year. What happened?
DH: Fraud and viewability are industry wide issues inherent to the supply side of the equation. As a supply player one can either originate audiences or aggregate audiences. The vast majority of Blinkx supply comes through aggregation, meaning that we are at least one step removed from the consumer, and as a result our policy is “don’t trust, verify”.
However, just to clarify, the landscape is complicated and can be split into three areas of verification, including:
- Time in view
- Bot traffic
- White space injection traffic
Currently there is no single company in the market that deals adequately with all of the above areas. For example, Forensiq is a pure fraud verification specialist, whilst Nielsen focuses on audience verification. Each party has its strengths and weaknesses and, as a result, must be used in tandem to ensure that all traffic is clean and safe.
But, as with most things, there are costs attached to all of these tools. To ensure optimal results and put the minds of advertisers at ease, we have partnered with the leading industry providers across the areas noted above, primarily Integral Ad Science, Nielsen, Forensiq and Moat.
In order to ensure the market reaches the level of confidence required, it will take all the businesses and brands within the video ecosystem working together to combat the issues that challenge the supply chain. Video DSPs, Exchanges and supply side players need to pool their resources to ensure we are collectively providing a quality advertising environment. The role of identifying, tracking and removing this type of traffic should be our joint responsibility.
Having verification companies that work just for the publisher or the exchange or the supply side is not enough to clean the market, collective partnership is the way forward.
Could you explain how you have been working with Blinkx? Can advertisers now trust the inventory?
ES: We’re delighted to partner with blinkx and feel good that they have a command of the tools that eradicate fraud. They are now using a series of best-of-breed verification tools ranging from viewability to brand safety as well as our fraud detection tool, and together, I believe that blinkx is taking concrete and thorough action to bolster its commitment to provide advertisers with quality supply.
Forensiq and blinkx are working together in a genuine partnership, and we provided our technology, totally free, for a trial so that blinkx could ascertain the value and veracity of our analysis prior to making any sort of commercial agreement. It’s how we work with all our clients. Our data shows true actionable value – in real time – which provides blinkx with meaningful differentiation around fraud detection and remediation within the marketplace.
Diligence needs to be maintained as different, more sophisticated strategies emerge from the criminal elements that drive fraud. We feel that together, blinkx and Forensiq have implemented a solution that truly leads the way in real-time fraud detection and blocking and we will continue to do so.
One of the things Edelman expressed concerns about in his blog post was the traffic coming from Zango and AdOn, both of which were acquired by Blinkx. Was there a problem with this traffic and is Blinkx still using these technologies for traffic generation?
DH: blinkx did not acquire Zango. It purchased certain assets in bankruptcy. However, blinkx does operate an ad supported software business that drives quality conversion traffic. This supply is generally not suitable for video advertising and is therefore not part of our video supply pool in the UK. AdOn is an asset that is expert in screening, filtering and optimizing supply. However, it is one of many supply sources in our supply arsenal but is not used currently internally to drive audience.
blinkx distributes professional content from over 1,000 providers across the web through its video platform. Our platform is divided into 16 interest channels ranging from news and politics to sports, fashion, lifestyle, tech, gaming and more. To distribute the best quality content we align ourselves with the likes of Reuters, Bloomberg, Condé Nast, National Geographic and other leading content producers. Our role is to ensure that their content is viewed and managed in the best way possible ensuring its integrity, the integrity of the audience and the integrity of the advertisers.
While Blinkx and Rocket Fuel have been singled out for bot traffic in the media, many say that this in fact an industry-wide problem? Is that true in your opinion?
ES: It is absolutely an industry wide problem. There is no part of this ecosystem that should view the eradication of fraud from digital channels as a benign issue. Agencies, publishers, media owners, intermediaries, ad technology vendors, analytics platforms; these should all be employing fraudulent activity screening into their strategies to ensure that fraud has no back door into the channels we use daily to drive value to advertisers.
How have advertisers responded to your revamped verification and fraud detection strategy?
DH: Thus far clients and partners are pleased with our actions. The most important thing now is to communicate the message that we are taking a proactive approach to this issue and that we will continue to lead the industry in tackling this moving forward. To ensure optimal success, the industry must move forward together.
Sometimes publishers find they receive fake traffic when they’ve no relationship with the bot. Why do bots sometimes generate fake traffic on sites they don’t profit from?
ES: In my experience, bots used for the propagation of fraud are very rarely used where there is no direct profit to be had. I think it is important to remember that bots are not a generic problem; there are good bots and bad bots, so to blindly label all bot activity as suspicious is inaccurate. However, we can deduce from our analysis, the situations where bots and networked bot systems are being used to drive fraud and to mimic human behaviour rather than as legitimate activity.
Fake impressions, clicks and conversions can sometimes be found operating behind-the-scenes on some very prominent and premium online properties, with no awareness on behalf of these property owners that this is occurring. So we should not assume that publishers who are found to have fake traffic are always culpable. That is exactly why we feel the eradication of fraudulent traffic is an issue for all and not just a buy-side problem in isolation.