August 9, 2018 — By: Tom Grubisich, Street Fight
This Q&A initially appeared in Street Fight. Click here to read the original.
Daily newspapers and other local news providers continue to be among the major producers of fraudulent traffic in programmatic advertising.
According to new research from Pixalate, a cross-channel fraud intelligence company that works with brands and platforms to prevent ad fraud and improve ad inventory quality, about a quarter of all smartphone app video and smartphone app display activity is “invalid traffic” (the technical term for what is largely fraud). Much of the fraud—up to 40%, based on “share of voice” that advertisers buy to get exposure—is on news sites, including local ones.
As programmatic becomes the dominant form of digital advertising—it should hit $65.6 billion in display alone this year, according to eMarketer—fraud becomes an ever-more-pressing issue for news publishers as they seek to win over businesses, who have been spending increasing shares of their messaging dollars with Facebook and Google.
Pixalate’s new research on ad “viewabilitiy”—which includes ads that are “at least 50% on screen for one second”—isn’t great news for news publishers either. In the research, “News & Media” sites rank in the higher ranges with about 57% viewability for desktops and 52% for smartphones, but that performance still means that close to half of ads on news sites don’t meet viewability standards.
In this Q&A, Pixalate’s Amy King, VP of product marketing, talks about the implications of the high rate of fraud in programmatic ad traffic and what publishers should be doing but often aren’t to lower the bad numbers.
Pixalate’s recent quarterly report on programmatic ad fraud says: “It’s a widespread problem that the industry needs to get serious about and fast.” Do publishers have the primary responsibility to act, and do you see them taking steps to do so?
Every stakeholder in the programmatic ecosystem should take responsibility in the fight against ad fraud. Many publishers, and particularly local publishers, may not have the infrastructure needed to do the heavy lifting themselves, but they can and should demand it from their vendors.
Publishers should make sure that all third parties that have access to their pages are doing due diligence in combating fraud. A lot of publishers have not been as proactive and consistent in this regard as would be ideal. Publishers should consider balancing the number of vendors they use to monetize inventory against their ability to conduct reasonable diligence regarding such vendors. By mitigating (and vetting) the number of companies that have access to their inventory, publishers can improve security and reduce the potential for ad fraud.
How did local news get into its predicament with programmatic ad fraud?
At the outset, programmatic offered local news publishers the opportunity to sell their inventory without the high cost of direct sales. However, as the programmatic ecosystem expanded, the number of companies involved in ultimately showing an ad to a consumer on their website grew as well.
The more companies present on the site in some form (tag, pixel, ad chain, etc.), the more opportunities for fraud. Further, because programmatic is an auction system, it was financially advantageous for local news publishers to allow many companies to both offer and bid on their inventory. This created a situation where many local publishers had a great number of companies with some access to their sites and data, but little transparency into who might be, perhaps unknowingly, bringing along a fraudulent intender.
Mobile app videos and display have higher rates of fraud activity compared to other devices and channels. Why so?
Fraudsters know that marketers are spending most of their money on mobile advertising, specifically mobile in-app, and they follow the money. It’s challenging for marketers to combat because they have a lack of visibility into in-app advertising because so few companies are able to detect, monitor, and prevent ad fraud in-app. Additionally, in-app video advertising is the top choice for app install advertising and therefore most prone to fraud activity.
Focusing on app videos on Android and Apple devices, why is fraud so high in some categories and so low in others?
There is a wolf-in-sheep’s-clothing phenomenon happening in the mobile app world. Ad fraudsters know which app categories are popular among real users, and they can use that knowledge to their advantage by creating seemingly legitimate apps within popular categories.
Unbeknownst to the users, their phones can be turned into vectors for ad fraud. We highlighted a recent example of this with the MegaCast-Chromecast Player application, which had been downloaded by over 1 million Android users.
One of the most susceptible categories for in-app video ad fraud on Android devices is the “Tools” category, which had an invalid traffic rate of 34.6% on Android smartphones in the time period we studied. These include flashlight apps, “security” apps, and other apps meant to improve or enhance the usability of the phone. On iOS devices, for video advertising, the “Weather” category was hit hardest by fraudsters in the time period we studied (40.2% rate of fraud activity).
Fraudsters are exploiting the verticals that users are selecting to protect or enhance their devices. Another dangerous and exploited facet is the extraordinary permissions that certain types of malicious apps are granted to operate. Many users do not give much thought to the number of permissions they give an app, especially if the app is supposedly going to help with their security, and fraudsters are taking advantage of this fact. Once gone rogue, the fraudsters control the device and can abuse it, even operating it while it is supposedly in “sleep” mode.
Why is “News & Media” so dominant in “share of voice” among desktop categories of content that advertisers buy to get the exposure they want?
Around 40% of desktop display and desktop video programmatic advertisements went to domains in the “News & Media” category in the time period we studied. It’s possible that the buildup to the midterm elections is fueling higher volume in this category.
Why is it so difficult to identify fraud intenders and bar them from the programmatic process?
There is significant financial incentive and an increasing number of ways to execute ad fraud, and few entities have both the resources and incentive to go after them. This is starting to change as advertisers demand greater transparency and demonstrable solutions. Ad fraud is an ongoing challenge for marketers, as well as for industry and governmental bodies tasked with enforcement and oversight, and constant monitoring is required to stay on top of new threats.
What does the “ads.cert” tool do to stop fraud more effectively than the older, more limited “ads.txt” tool?
We shared a blog post in which we explain how ads.cert can help reduce ad fraud in programmatic. Pixalate fully supports the IAB’s ads.cert initiative, which remains in a period of public comment.
The Local Media Consortium, which represents 75 newspaper companies, broadcasters, and digital media companies that have 1,700 digital news sites, says it has a strong commitment to fight fraud and has recommended that its members take specific action to fight it. But the LMC can’t order its members to take such action. Is this a problem? Has the LMC worked with Pixalate in any way?
Pixalate has not worked directly with the Local Media Consortium, however, we do work with a variety of organizations that seek to help their members reduce ad fraud. We find that the better companies understand the problem, the greater their ability to enlist vendors that are best equipped to solve their ad-fraud challenges.
Sign up for our blog to stay updated with new stats, trends, and analysis on digital ad fraud.