The premise is simple - advertisers pay money to have their ads seen. They make a purchase and expect results. Without these simple transactions, the advertising ecosystem wouldn’t exist.
But sometimes, it’s not that simple. Ad tech and media companies live and die by their ability to give advertisers what they pay for, wholeheartedly, with no misgivings. However, there’s an evil enemy lurking around nearly every corner in the industry – fraudulent traffic or bots. In order to give advertisers what they want, an impression must be viewable by a human consumer, but fraudulent traffic taints those impressions, leaving the customer (aka the advertiser) with buyer’s remorse.
Fighting Fraud is a Commitment - It All Starts with Data
Fighting fraud is much like your favorite crime drama. It all starts with a crime scene (the Internet) where even the smallest detail may represent evidence (data) of a crime (non-human traffic consuming impressions).
If data is the foundation, the multitude of potential clues that can identify suspicious activity, then data + sciences represent the ability to separate the evidence from the noise. Using its team of data scientists to model audience composition, you can create custom targeting based on analysis of demographic data, and develop industry-leading research on the efficacy of video advertising. Together with engineering, data scientists can track hundreds of signals that reveal potentially fraudulent activity.
Validate with Forensics
To continue the crime scene analogy, once the CSI team has uncovered every last piece of evidence tied to a crime, the wise-cracking, street-smart detective puts the puzzle together. This is the role of the traffic quality forensic investigation team. These cyber-detectives take the results of data + sciences, use their expertise sifting through log files, and peel the layers of the onion away to determine whether suspicious activity is truly nefarious, or if the traffic has legitimate origins that can be explained. Moreover, the result of their investigations, the learnings and revealing data, are incorporated into the models created by our data scientists to better detect future malicious activity.
It’s critical to ensure that new supply entering the ecosystem brings humans, and not bots, into the network. At YuMe, we not only review new publishers to make sure they meet content, legal and technical standards, but we also look at a site’s reputation. A new site may be well formatted and have a professional design, but we dig deeper if our research determines if there is a high risk that this new site will bring malicious traffic into the network. It’s very much like dusting for fingerprints at a crime scene – there’s more that can be seen by the naked eye.
Though the primary objective in mitigating invalid traffic or non-human activity is to protect advertisers so they get what they pay for, it’s critical to consider the health of the larger video advertising ecosystem in which companies like YuMe operate. Our advertisers interact with many parts of the ecosystem; they work with other networks, exchanges and publishers directly to obtain the largest possible reach for their marketing campaigns. When we find a highly malicious and extraordinarily damaging fraud scheme in our network that has a broader impact in the ecosystem, we will share the intelligence that will help other parties stop this activity in its tracks. Not only does this protect our advertisers wherever they spend their marketing dollars, it’s simply the right and ethical thing to do.
Ad tech companies can sit on their secrets of combating ad fraud, but who does that really help? No one. By sharing knowledge of best practices and collaborating on research, everyone benefits – not just those with a pre-existing relationship with a specific technology vendor.
It’s About the Advertiser
Ad fraud is a constant, never-ending battle, requiring constant vigilance. Ad fraud hackers are smart – very smart – and they follow the money, meaning they will continue to try to exploit the video ecosystem, likely redoubling their efforts as more money shifts to online video. They are becoming increasingly more sophisticated, but while that happens, ad tech companies need to band together to pool their collective knowledge and find improved, ever-evolving ways to stop fraud in its tracks. In order to ensure network quality, YuMe developed a Traffic Quality Dashboard – a heat map that tracks instances of online ad fraud across the globe.
Given the sophistication we have, and the broad swath of inventory we have access to, we see things individual publishers or smaller networks don’t. As a result of this, we see the value in educating publishers on what to look for and how to best behave so they can continue to get great brands running on their content; however, our main interest here is the advertiser. The customer. So, it is our job as part of the larger advertising community to make sure our customer is always taken care of.