We recently sat down with Matt McLaughlin, COO at DoubleVerify to discuss the current state of fraud in the industry. Given that DV is a proven market innovator with technology that delivers transparency and accountability, ensuring fraud protection and brand safety, we knew he would have great insights to share with us. This week’s DoubleVerify Q&A will cover high level definitions and the specifics for how certain types of fraud, like URL masking, take place. Next week, we will continue the Q&A format and dig into trends and recommendations for combatting fraud.
Q: It seems today that fraud is an umbrella term for all inventory quality issues. How does DoubleVerify define fraud?
Digital ad fraud happens in the “who, where and how” ads are delivered.
The “where” refers to the environment in which ads are being served, and fraudulent practices are designed to disguise the URL where a digital ad will actually be served with one that would looks more attractive to an advertiser, especially in a real time programmatic bid environment. URL masking and Impression Laundering (when an objectionable site launders its traffic through a seemingly legitimate front site) are the most common types of fraud in this category.
The “how” refers to how an ad is actually rendered on the page. Hidden ads served in tiny 1×1 pixels, stacked ads that are delivered on top of one another, and other tactics that misrepresent how the ad appears (or doesn’t) on the page, are ways that fraudsters cheat online advertisers.
The “who” refers to the overwhelming volume of non-human ad impressions (bots) that are counted and paid for by online advertisers. Bot fraud, in both display and video, has become the most serious area of traffic fraud to infest the internet today.
Q: How does DoubleVerify define a bot? What makes up a bot?
DV defines bot fraud as a robotic machine imitating human browsing behavior with the intent to defraud advertisers. This typically occurs when malware is unknowingly downloaded on people’s computers, typically running in the background. In other words, normal users sometimes get viruses and adware on their computers and as a result, a bot is installed in the background. This is an important distinction because it makes them much more difficult to catch.
Q: How rampant is ad injection in the video ecosystem?
Ad injection is a growing problem because it’s so profitable and so difficult to detect. In addition, there’s real demand for the inventory even if the advertisers don’t know what they’re buying. Say an advertiser has been buying inventory on cnn.com for years at a $5 CPM, and all of a sudden they see cnn.com inventory available on the exchanges for $1 CPM. They think “wow, that’s a great deal!” and buy up as much as they can.
The inventory is in fact appearing on cnn.com, but it’s not CNN that is initiating the video impression or getting compensated for it. Instead it as an intermediary company that developed the adware, toolbar or extension that caused the ad to be served on the user’s machine. The result is that neither CNN nor the advertiser is the wiser, and the advertiser is not actually getting the native publisher-initiated ad impressions that were represented to them.
This is an obvious and severe problem for the ecosystem because publishers are effectively being stolen from and without a healthy environment for publishers to pay their bills, they won’t be able to produce the quality content that drives these visits in the first place.
Q: How does URL masking take place and how can I avoid masked inventory?
URL masking manifests itself for a few different reasons but the ultimate goal is to hide the location where the ad actually appeared. Most times, the sites engaged in this practice have real users but its content has low advertiser interest. The owners of these content sites, often pornography, gambling or copyright infringing websites, look for ways to monetize this user audience so they misrepresent the inventory as appearing on a domain with content more desirable to advertisers.
This can be accomplished through impression laundering or sometimes even just hardcoding the “bid domain” into their ad tags. The result is that the intermediary sellers, and subsequent buyers, of this inventory think it’s appearing on a brand-safe or premium domain when, in fact, the impression appears in an environment the advertiser never would have agreed to.
Because of the complexities and various manifestations of this problem, the only way for advertisers to avoid this is to employ technology that enables them to see the true domain that their ad appeared on. Blacklisting obviously won’t work by itself, because the domain of the masked site can change rapidly and re-appear in the ecosystem under a new domain to start the process again.
Read more from our Fraud Series:
- Fraud Series Part 1: The Irresistible Allure of Ad Dollars
- Fraud Series Part 2: Hidden Ads
- Fraud Series Part 3: Botnets & Hijacked Devices
- Fraud Series Part 4: Proxy Traffic
- Fraud Series Part 5: Domain Bait & Switch
- Fraud Series Part 6: Q&A with DoubleVerify, the Leading Industry Performance Innovator
- Fraud Series Part 7: Q&A with DoubleVerify, Top Trends & Recommendations