Last month, CNBC reporter Megan Graham tried to find a post of hers that had just gone live only to discover that the “New York Times Post” had published a similar story with the same headline. After clicking through, she discovered it wasn’t just the same headline, but her whole article.
However, when she looked at the site, which was filled with plagiarized content, she noticed it was displaying ads from legitimate companies such as MetLife and Neustar Security.
This got her curious: Just how easy was it to set up a plagiarism site and rake in some unearned ad revenue.
After buying a domain and setting up shop on a cheap WordPress hosting account, Graham set about applying to as many advertisers as she could. Though many did reject her for having too little traffic or because she didn’t make enough money.
Others, however, accepted her very willingly. One company, Sovrn, actually spotted that she was publishing plagiarized content and initially denied her before approving it 24 hours later for no clear reason.
Google, for its part, did reject her due to her scraped content, but still ended up in her ads.txt after services that partner with Google did accept her.
Soon enough, her fake news site was showing ads prominently for companies like Chewy.com, Overstock and Wayfair and more. Countless legitimate companies advertising heavily on a fully (and openly) plagiarized website.
But, while it would be easy to put the blame on a few bad actors or a bug in the system, the simple truth is that the entire internet advertising industry is built.
Middlemen Upon Middlemen Upon Middlemen
A study by the Incorporated Society of British Advertisers and PwC, which was also cited in Graham’s original article, highlights just how opaque and obtuse the internet advertising industry can be.
According to the report, the 15 advertisers they looked at had nearly 300 different supply chains to place ads on 12 different publishers. This means that, if an advertiser wanted to put an ad on a specific site, there were likely dozens of different ways to do it, each involving layers upon layers of middlemen.
The result of this is that, for every dollar that advertisers spend on an ad, just 51 cents reaches the publisher. The rest is eaten up by an assortment of middlemen including the advertiser’s agency, both demand and supply-side services as well as technology companies working from both sides.
Perhaps most worrying is that, despite the study’s rigor, it could not account for some 15% of the money. This “unknown delta” made up a larger percentage of the money than any other actor besides the publisher.
The result of this is that an ad takes a large number of steps between the advertiser and the publisher. In many cases, an advertiser will hire an agency that will then work with demand-side providers, who in turn work with technology providers that then reach out to supply-side providers, who in turn work with tech providers to put ads on the publisher’s websites.
If an ad winds up on a plagiarized website, the advertiser likely hasn’t seen the site, doesn’t know the advertising network it’s on and has little idea of how it got there. Between the advertiser and the publisher sets a gaggle of middlemen, partners, technology services and other companies.
All of these layers exist as a means of better automating and streamlining internet advertising. They aim to help advertisers get their ads in front of the best audience possible for the best price while helping publishers maximize their profits.
The problem is that all of these layers also make it easy for bad actors to slip in and make it difficult to do anything about it.
That, in turn, is exactly what Graham demonstrated with her experiment.
The Root of the Problem
The problem is straightforward. Advertisers don’t want their ads on sites like “The New York Times Post” and every dime that goes to a plagiarist’s website is one taken from legitimate publishers. Neither side wants this.
But the problem is that if just one company in those nearly 300 supply chains is less-than-careful about their vetting or has lower standards, bad actors can slip in and infect the entire process. There’s not much that anyone at either end of the chain can do because those bad actors are buried deep in the stack of middlemen that sit between them.
Advertising companies insist that such bad actors are usually caught fairly quickly. However, that’s pretty much meaningless as there are no long-term consequences for them. The spammers themselves just try again because, as Graham showed, it doesn’t take much work to set up such a site. Meanwhile, the companies that opened the door for them retain their partnerships and connections.
No matter what checks or vetting most companies do, as long as one or two keep the door open, the plagiarists will always find a way inside.
Because of that, there are really only two solutions.
The first is what was proposed Bob Hoffman in Graham’s article: Advertisers should buy directly from publishers.
Yes, it would likely cost more for advertisers but they would have greater certainty in what they were purchasing. It would also mean that more of the advertising money actually made it to the publishers, doing more to support legitimate content creation.
The other is keeping some of the current structure, but implementing a broader consolidation and flattening of the advertising industry. While I’m typically skeptical about industry consolidation being a good thing, it’s clear this is an industry that’s shown it can not effectively police itself with its current structure. Many of the middlemen currently embedded likely are either not necessary or could have their services rolled into another company’s offerings.
However, this quickly delves into areas that I am not an expert in. Which companies are no necessary? Would industry regulation be more effective with a small group of companies? How would this consolidation happen?
I don’t know those answers. But what is clear is that, with an industry this layered and complicated, it’s proving impossible to block bad actors and keep them out and that risks harming the industry more broadly.
After all, a couple of plagiarists and fake news site may not be a big deal, but as stories like these grow, so does the backlash from all sides.
Ultimately, the real reason this problem persists is that there’s not enough motivation to fix it. While there are certainly some solutions that could work, the resolutions are seen as being worse than the issue itself.
When YouTube was called out for putting ads on terrorist videos, they began the “adpocalypse” and demonetized countless channels. The same held true when YouTube reached a $170 million settlement with the FTC for marketing to children illegally and then began to restrict monetization on channels targeted at kids.
Though YouTube is just one company, it’s a mammoth and complicated one. These stories show that the advertising industry can make major changes when motivated. However, it does show that those changes hurt and have impacts for everyone, including the innocent. Still, if the will were there, I have little doubt that the industry could solve this completely.
However, the will isn’t there. Whether it’s because the plagiarism problem isn’t that big or if it just doesn’t impact the bottom line enough, it’s not being touched.
Until such a time as it is, expect to see a lot more sites like “The New York Times Post” and other sites that scrape news stories and articles. There’s nothing really stopping them, but there’s a system ready to reward them.