New York, NY – July 16, 2014 – Integral Ad Science, the leading global provider of actionable advertising intelligence for buyers and sellers of digital media, today released its Q2 2014 Media Quality Report, updating the industry on important metrics – True Advertising Quality (TRAQ) Score, ad fraud, viewability, and brand safety. This quarter, the report features additional verticalized data, comparing how certain verticals’ digital ad campaigns performed. It also revealed no significant change in performance by channel as the industry has mostly stabilized through the consistent adoption of technology to keep up with the challenges that the digital advertising industry is facing.
Integral’s quarterly report relies on robust data processed through its industry wide integrations, including the top five demand side platforms (DSPs), nearly all of the major ad exchanges, all agency holding companies, and more than 70 of the “Advertising Age Top 100 Advertisers.” Each impression is evaluated based on environment and exposure components and receives a TRAQ score, a holistic measure of media quality. By including attributes such as fraud, viewability, ad clutter, brand safety, and professionalism, TRAQ provides an all-in-one snapshot of media quality, allowing buyers and sellers to understand the true value of their inventory.
In addition to Integral’s standard media quality metrics, the verticalized data included in the report indicates technology companies suffer from a larger amount of fraud (17 percent) than other verticals such as consumer packaged goods and quick service restaurants (from 6 to 6.5 percent). The significant difference between the verticals can be due to the difference in media buying preferences, as higher fraud rates are found on the exchanges (16.5 percent) and ad networks (10.5 percent) as opposed to direct publishers (3.5 percent).
Through its continuous study of digital fraud and botnet behavior, Integral’s data science team has observed new tactics over the past quarter. To better camouflage their activities, bots are ordered to visit a mix of non-fraudulent sites. By doing so, they are able to populate cookies and become more attractive for targeting. This tactic confuses bot detection models that rely heavily on detecting infected machines predominantly visiting sites known to be fraudulent. To mask their operations, bots were also observed recycling new domains more frequently than before, in order to escape simple detection models labeling users visiting ‘fraudulent’ sites as bots. Lastly, the data science team is reporting botnet attempts to cover up their tracks by increasingly using browsers posing as Chrome, which many detection products have limited ability to recognize.
Despite the recent lift of the MRC advisory against transacting on viewable impressions, Integral found no significant change in viewability. This is expected to change in the upcoming quarters, as the industry adjusts and adopts new expectations. For maximum see-through rates, Integral uses both page geometry and browser optimization methods, and excludes ad fraud when measuring and reporting on viewability.
Although no significant change was found in brand safety levels, when broken down by category, the report reveals that risky impressions most often land on adult content (41.8 percent). This is indicative of the volume of adult content on the Web and the traffic it receives. In addition, the 13.4 percent of impressions flagged on illegal download sites is notable as it highlights the need for continued vigilance against piracy.
“As the leader in the category, Integral Ad Science is the only vendor to have access to ‘bid data’ though its partnerships and industry-wide integrations,” said Scott Knoll, CEO of Integral Ad Science. “As a result, our data offers an accurate view of media quality trends that are needed to understand the overall direction of the industry.”
To view the full report, click here.