October 9, 2014

Upping the Ante on NHT

Paul Barford Paul Barford
SVP, Chief Scientist

Over the past several years, comScore’s Brian Pugh has been outspoken about the problem of non-human traffic (NHT) in online advertising, describing the nature and scope of the threats and outlining the steps needed to eradicate NHT from the ecosystem. Fortunately, due to the efforts of Brian and others, the NHT issue has received growing attention in the popular press and with IAB and MRC. Wider industry recognition of the problem is a critical step toward the broad adoption and use of policies and mechanisms that can reduce NHT.

There are many challenges in developing effective countermeasures for NHT. First and foremost, since there are significant financial implications, it must be assumed that the parties that are engaged in NHT generation and other forms of invalid online advertising are highly motivated, intelligent, capable, and well-equipped. They cannot be taken lightly or underestimated, and the result is a threat landscape that is broad and evolving rapidly.  To that end, developing and maintaining effective countermeasures requires superior technology and data assets and an experienced team committed to addressing threats in a comprehensive fashion over the long term.

Recently, comScore doubled down on its commitment to dealing with NHT by acquiring MdotLabs, a Madison, WI-based company that I co-founded in 2013 with Timur Yarnall to develop comprehensive NHT detection and mitigation solutions. Timur’s experience in online advertising and my background in large-scale Internet data analytics and security provided a combined skill set essential to addressing the current and looming NHT threats. Our high-level approach to addressing these threats is to develop methods for searching through web browsing data to identify non-human behavior. Analyzing web browsing data requires significant capabilities in big data analytics, as well as a detailed understanding of how humans behave and respond to various marketing stimuli.  

With these principles in mind, the Mdot team developed a sophisticated NHT detection and mitigation solution called AdSecure. At the core of AdSecure is a set of machine learning, signal processing and statistical algorithms that detect a variety of NHT threats such as bots, human view/click farms, non-bot tools for traffic generation (e.g., simple scripts, plugins, pay-per-view networks, etc.) and mechanisms for hiding, redirecting or obfuscating ads. We have been continuously evolving these methods to address new threats in display, video and mobile advertising.

By joining forces with comScore, we now have an opportunity to extend our NHT threat detection and mitigation capabilities even further. In addition to Mdot’s existing honeypot system, we now have data from comScore’s two million person global panel and data from billions of ad impressions that are captured on a daily basis. These unique data assets bolster our view of both people and pixels, thereby providing even more angles from which to detect NHT and improve our ability to rapidly identify emerging threats. We also appreciate the opportunity to expand the use of our NHT detection technology by integrating into comScore vCE and vME, which are used by hundreds of media buyers and sellers.

We are progressing quickly in this integration and look forward to bringing the digital ad ecosystem NHT detection that is second to none.

Learn more about comScore NHT triple detection.