Fraud continues to be one of the greatest concerns in the digital advertising space. Despite the progress that has been made, our industry evolves at a rapid pace, which led to the emergence of new deceptive practices. As a result, buyers still have quality concerns around due to the presence of botnets, pixel stuffing, domain spoofing, and ad stacking – let alone the complex issues of spurious content.
Over the past year, buyers, DSPs, and some exchange platforms called for the industry to start policing itself. In fact, as recently as December of 2015 a study from the Interactive Advertising Bureau (IAB) found that advertising fraud cost advertisers $8.2 billion per year—with $4.2 billion lost to “nonhuman traffic,” $1.1 billion lost to “malvertising-related activities” and the remaining $2.4 billion lost due to “infringed content.” It’s no surprise then that the buy-side of the industry was up at arms over these issues.
As tools became available from companies like Integral Ad Science (IAS), ForensIQ, White Ops, MOAT and various others, SSPs and publishers began scanning their inventory and removing the majority of suspicious and malicious impression requests.
These efforts provided industry-wide benefits. As inventory is scanned and turns out to be quality rather than fraud, buyers’ trust in the publishers in question inevitably increases. This trust in turn results in demand partners increasing their willingness to buy a publisher’s inventory, as they can be more secure in how their campaigns are targeted and delivered.
Despite initial pushback from publishers due to the increased cost and the complexity of scanning and qualifying their inventory, we are seeing more publishers begin to comply with buyers’ demands as proof that advertiser spend will shift away from questionable inventory to verified inventory continues to stack up.
Fraud and Quality: Not One-Way Streets
While the sell-side of the industry made large strides in verifying the quality of their inventory in response to requests from the demand side of the industry, these actions have been rather one-sided to date. Malware, spoofed landing pages, unidentified creative attributes and various other nefarious things continue to flow forth from the buy-side.
Malware and other fraudulent activity is a danger to all parties within digital advertising, with cascading effects. Not only can a user’s device become infected via malware, for example, they also are likely to install an ad blocker, or to avoid a publisher’s site entirely as a result of other fraudulent activities. This causes a reduction in inventory volume, which can in turn reduce not only a publisher’s revenue, but also their trust in programmatic and potentially lead them to move back towards ‘safer’ direct sales models.
The issues compound as SSPs and exchanges are forced to build out additional tools to throttle back DSPs in an effort to maintain quality on their platforms based on aggregated advertiser data. Unscrupulous advertisers are running campaigns that knowingly disrupt the user experience or spoof landing pages in an attempt to fool publisher-implemented block lists. In the end, quality advertisers can get caught in the slurry, losing access to high-value inventory because of the actions of other, nefarious buyers that happen to be utilizing the same platforms.
At the same time, while the OpenRTB standard the majority of buyers use allows creative attributes to be passed to SSPs, they are almost universally not being passed. The OpenRTB standard also allows for SSPs to provide creative restrictions to DSPs and buyers, but this information is passed it’s often ignored. As a result, SSPs were forced to develop their own solutions in order to protect their publishers from creative formats that are likely to interfere with the way in which publishers intend their content to be consumed (one of the greatest sins publishers can commit in the eyes of consumers, who have little to no tolerance for interruptions to their user experience).
At PubMatic, we not only developed our own solutions, but also partnered with outside companies to handle Abhorrent Advertiser Activity, and we’re not alone in doing so. We, and other platforms, employ teams of people dedicated to finding and stopping ‘bad ads’. But these are just Band-Aids that treat the symptoms of the problem rather than addressing the real causes. While we can pre-scan demand to ensure compliance, we must both identify and classify before we can perform any comparison. Until we treat the source of these issues, we will remain in a reactive state.
A Long-Term Solution is Needed—and Does Exist
Just as SSPs and publishers implemented tools and established processes to police inventory, DSPs and buyers must do the same and help to eliminate questionable advertising. While the tools to achieve this exist today, we need to ensure that our technologies, processes and policies allow us to take advantage of them. Self-serve platforms must implement a layer of creative approval. Buyers and advertisers must register their creative attributes, for all creatives, as well as the actual landing page, and the advertiser name needs to be passed in all bid responses. These three reasonable actions, all of which are supported by the existing OpenRTB protocol, are all that is required to show good faith in the eyes of the publisher community.
An open, fair and transparent market requires that both sides see exactly what is being bought and sold. Publishers and SSPs have done their part, and now we ask the same from the demand side.
We urge those companies on the buy-side of our industry to take steps towards advertiser transparency under the banner of publisher confidence. Programmatic buying, in all its forms, is at risk whenever malware or an explicitly forbidden creative is served. While no solution will be 100% effective, any effort from the demand side to address the root of the problem will have positive effects. The inclusion of creative attributes and correct, transparent, landing page URLs will both reduce brand control violations and provide additional targeting data to buyers to help their own spending decisions.
At PubMatic, we continue to improve both our internal processes as well as our products and solutions, while also partnering with the best industry solutions for identifying and counteracting malware and other fraudulent activity. We continue to strive to provide the highest quality demand while abiding by our publishers’ creative restrictions. In this ever-evolving world of digital advertising we will remain vigilant. We ask that you do the same.