Login is restricted to DCN Publisher Members. If you are a DCN Member and don't have an account, register here.

Digital Content Next


InContext / An inside look at the business of digital content

Three inventory quality trends to watch in digital advertising

July 24, 2018 | By Eric Bozinny Director, Inventory Quality—PubMatic @pubmatic

Ensuring the high-quality of inventory across the PubMatic platform, as it flows from sellers to buyers, requires strong policy, which standardizes compliance enforcement and operational coordination across account teams so that we spot issues early and often. It also requires a strategic focus on identifying what lies ahead for quality.

I’d like to share my thoughts on a few growing trends I expect to see regarding inventory quality. These views come from an amalgamation of inputs: my 10 years of experience in managing inventory quality, the signals and other clues arising from my daily quality operational work, and deep-dive investigations. Buyers have shifted their emphasis to quality and are focused on working with other quality-centric professionals across the industry.

Here are three major inventory quality trends:

Over-Reliance on Fraud Detection Technology

The industry has clearly spoken – third-party fraud detection is now considered “table stakes” for any large player, buyer or seller, in the digital advertising ecosystem. Though fraud detection vendors play an important role in helping to identify and avoid invalid traffic (IVT), I would advise treating this service as one tool among others to help improve quality.


No vendor measures quality the same way, yet many of them share the same MRC certification for Sophisticated IVT (SIVT). PubMatic uses a combination of IAS and White Ops to monitor invalid traffic rates and identify problematic pockets of inventory. However, many buyers use different fraud detection vendors and they may report very different results for the same inventory. These variances can be explained by unique proprietary methodology and differences in sampling where one vendor may look at a completely different subset of the same inventory vs. another vendor.

For example, if “Buyer A” is reporting their inventory is 100% IVT, while PubMatic’s White Ops reporting is showing 1%, the promise of fraud detection technology as a standalone method to identify non-human traffic breaks down. Yet, when stepping back and viewing fraud detection as a starting point, conflicts between buyers and sellers are more likely to come to an agreement. I believe this because both parties can recognize that even with big differences in fraud reporting, a deep-dive investigation will uncover other signals that likely support one report or the other.

In this specific example, I may find other evidence supporting the buyer’s claim and could come to a mutually agreeable conclusion (e.g. refund, blacklisting, termination of supplier, etc.). However, as often as not, my investigation might raise no other red flags to indicate poor inventory, and thus I would propose limiting access to that inventory for this buyer.

Growing Importance of Content and Audience

GDPR has impacted the ability of marketers to fully utilize the targeting potential of cookies and audience profiles in EMEA due to the regulatory changes in consent and privacy. One could argue GDPR is a direct consequence of the rise of ad tech and the lack of self-regulation concerning how the data of consumers are used in the targeting of advertising. Therefore, it is not unlikely that similar consumer data privacy and consent laws will spread around the globe. This will further reduce the efficacy of cookies and precipitating the return of contextual targeting for online advertising.

What does this mean? An increased focus on the quality of content—addressing fake news and brand safety concerns—as well as the quality of the audiences who consume this content as important quality trends in the marketplace.

Recognizing the importance of content and audience to buyers, PubMatic evaluates domains and apps not only on the level of IVT but also on the value of the audience and originality of the content. For instance, an organic, loyal audience is preferred to consumers acquired from other sources. We also avoid content farms and look-a-like sites that exist only as a necessary backdrop to sell ad impressions.

The Importance of Whitelisting and Ads.txt

Since ads.txt was introduced in 2017, adoption has been accelerating. Pixalate reports that as of March 2018, more than 50% of the top 5,000 programmatic sites have adopted ads.txt.

While ads.txt is a valuable tool to combat domain spoofing, it provides no inherent protection against IVT (bots and fraud). Further, it does not guarantee the quality of a domain’s content and audience. For example, a domain created solely for the purpose of driving bot and/or acquired traffic through pages filled with content stripped from other sources can have an ads.txt file, but still be a bad source of inventory.

Thus, working with a whitelist of trusted domains is the single best practice for both buyers and resellers of inventory. Being familiar with the domains on which ads are running, and avoiding all other domains, is the best prevention.

I strongly suspect that most of the behavior leading to poor quality inventory comes from the point where money changes hands—the domains and apps where advertising is consumed. By wisely choosing which domains to work with and working with only whitelisted domains, many quality issues will be avoided entirely. Alternatively, when a small group of domains isn’t enough to meet inventory requirements, working with trusted partners can also provide improved inventory quality and improved brand safety.

Liked this article?

Subscribe to the InContext newsletter to get insights like this delivered to your inbox every week.