• by George Simpson , Featured Contributor, December 17, 2020

Just as the Interactive Advertising Bureau Tech Lab was launching a new program to fight fraud and increase transparency in the connected TV business, Moat, a division of Oracle Data Cloud, said it had detected a scam that generated fake video ad impressions worth about $14.5 million. The fraud took advantage of the programmatic ad buying/serving that is fast being imported from digital to other platforms.

Some claim that buying inventory directly from publishers or trusted partners can avoid this type of fraud (although it tends to be 10 to 20 times more expensive.)

Ad exchanges like Yahoo’s Right Media, Google AdEX, Microsoft AdECN, AdMeld, PubMatic, The Rubicon Project, and Open X all started building RTB software that worked as a demand-side platform, a supply-side platform, an ad exchange, or some combination of the three between 2007 and 2010, so programmatic is not some shiny new object in the ad business.

Premium inventory is still mostly handled manually hence the higher cost, but remnant (long tail) is auctioned off to the highest bidders. Initially, programmatic demand was dominated by direct-response advertisers and performance marketing, but now is mainstream. On average, programmatic ads cost around $0.50 to $2 CPM, but are subject to ad blocking and the kinds of fraud uncovered by Moat.

By about 2014, programmatic was adopted by mobile ad sellers tracking users with SDKs and APIs built into mobile applications, but plagued with hidden ads, domain spoofing, click spamming, fake app installs, and click injection (when a new app is installed, the app/code sends a series of clicks to the MMP before the install is completed in order to get credit for the install).

To protect against programmatic fraud requires a long list of prophylactic activity like industrywide collaboration, standardization that establishes guidelines for better processes, ad formats, regulation policies, and best practices, constant monitoring, third-party verification and regulations to assure advertisers and publishers are on the same page.

The World Federation of Advertisers predicts ad fraud could cost advertisers $50 billion by 2025, yet nearly everyone in the programmatic business is bullish that it will continue to grow and eventually be used by nearly all media.

Why? Surely the savings generated by programmatic sales and execution are plenty offset by fraud and the need to hire third-party vendors to try and authenticate your traffic. So why would you want to invite that fraud into your industry if you are CTV, or any other kind of TV?

I am often told that advertisers want programmatic since it cuts down on their costs, including being able to use the same creative across multiple platforms, and gives the illusion of precision targeting. Publishers like it because they can tout their refined targeting capabilities and can get at least some money for inventory they could never sell directly. But that is partly a function of the infinite size of the internet -- not a big problem in the TV world.

As what is viewed on the flatscreen moves well beyond linear TV, the amount of ad inventory is growing mightily -- and with it, programmatic buying and selling. But it is worth the compromised trust incurred by revelations of fraud?