There are partners across the digital supply chain—brands, agencies, adtech vendors, publishers—that are deliberately profiting from fraud, and by extension, deliberately supporting organised crime.
That money then makes its way into other forms of crime, including drugs and prostitution. That’s according to a comprehensive 2018 study on cybercriminals by criminologist Michael McGuire on behalf of Bromium, which found that 40% of cybercriminals spend their money either reinvesting in cybercrime or on drugs and sex. Just short of one-third (30%) invest into value holdings such as property, art and wine, while 15% spend their money on lavish items like expensive jewellery to attain status or to impress others. So brand money is at best buying lamborghinis for Russian mobsters, and at worse, much darker things.
There are also partners within advertising that are doing their utmost to stop fraud, even if it is to the detriment of their business, and others caught somewhere in the middle, not fully trying to uncover the problem. The full spectrum exists. As with many issues in advertising, nothing is black and white.
“It's hard to tell sometimes where the line between being compromised or complicit is, and that has always been a challenge in this industry,” says Tamer Hassan, the chief executive of cybersecurity and ad verification firm White Ops.
“There doesn't seem to be a great deal of enthusiasm for brands to open the kimono and say, ‘guess what, 50% of our budget last year was actually fraudulent traffic’. And that flows through the agencies as well. They don’t want to put their hands up and tell their clients that a third of their budget went to fraud."
—Zachary King, MediaMath
“We've seen legitimate platforms in the ecosystem that are completely on the side of intentionally doing fraud, we've seen platforms and publishers that don't understand the problem and are more compromised, and we've seen those that don't want to fix it because it might impact their numbers."
Not everyone agrees with the concept that some in the ad industry are intentionally profiting from fraud, but industry experts can agree that the current business model of advertising is not incentivised to remove fraud completely. Or as MediaMath’s Asia VP of commercial Zachary King politely puts it, there can be a “lack of enthusiasm” within the industry to uncover the full extent of fraud.
“There doesn't seem to be a great deal of enthusiasm for brands to open the kimono and say, ‘guess what, 50% of our budget last year was actually fraudulent traffic’," King says. "And that flows through the agencies as well. They don’t want to put their hands up and tell their clients that a third of their budget went to fraud."
This attitude can stretch all the way to the top levels of the industry, King reveals. For example, a brand’s marketing team “will know they’re operating on the wrong metric”, but they have a KPI handed down to them from the C-suite, which may be using those metrics to prove business growth to investors.
“It goes all the way to the investment side—that's the kind of education level that it needs to happen at," King says. "I don't think that's going to change quickly or easily, and if you've got a CMO or CEO and they're there for short term, they've got to show growth.
“The advertiser will be the one at a loss because no one is seeing that ad, yet marketers are not moving off these KPIs. We need to review what does success look like for a brand."
—Laura Quigley, Integral Ad Science
“Conversations on ad fraud start at the technical level, but when you start to peel and unpack it it's so much deeper than that, it's more of operational.”
Ultimately, ad fraud doesn’t just increase revenue for fraudsters, but also for some players within the digital ad supply chain, Hassan surmises, and it’s time they understand where that money is going.
“Do we really understand what's on the other side of this?” Hassan poses. “It's not just a meaningless set of data, it's organised cyber crime that's funding the infection of users on the internet, it’s funding malware development.”
“As long as advertisers see this as a tax or as a data-quality problem, it will still be difficult to address. The reality is that ad fraud is a cyber security issue, not a viewability or measurement challenge,” Hassan adds.
But as much as perception is a challenge, it all boils down to metrics
Fraud has thrived so well in the advertising industry because the business model of advertising has always been driven by acquiring eyeballs, and as the old adage goes, money follows eyeballs (and fraud follows money).
“The driving force in business, media and advertising is capturing attention, but there should be a limited supply of attention on the internet, a limited supply of human eyeballs at any given moment," says Hassan. "Yet the reality is, if you ask for more and you pay for more, it is endless. Today we could put as many impressions on a blank web page with zero content as on the homepage of Campaign Asia-Pacific, without spending a dime on content and stories. And that is alarming."
There is no incentive for an advertiser to spend all of its media budget on fraudulent impressions, Shane Dewar, Essence's VP of media operations for APAC, rightly points out, “because that is not going to help their business at all”. Bots can’t make purchases, after all.
“I've been surprised by some of the conversations I've had with large brands where the moment you start impacting reach, it's actually a discussion of whether or not they should address it. Having people that are driven by these metrics—they're incentivised by fraud.”
—Tamer Hassan, WhiteOps
However, when a brand’s objective is volume of clicks "it is safe to assume that performance metric has more importance than the rate of fraud that is within that campaign,” Dewar adds.
Hassan says: “I've been surprised by some of the conversations I've had with large brands where the moment you start impacting reach, it's actually a discussion of whether or not they should address it. Having people that are driven by these metrics—they're incentivised by fraud.”
Laura Quigley, the Southeast Asia managing director of Integral Ad Science, continues: “The advertiser will be the one at a loss because no one is seeing that ad, yet marketers are not moving off these KPIs. We need to review what does success look like for a brand".
In many cases, not only do publishers and media buyers need to hit a certain amount of clicks to break even, but they can earn more money by exceeding these KPIs. Easy to see, then, why a person might be incentivised to turn to dubious schemes like traffic sourcing. This practice, which involves a publisher buying traffic to inflate the numbers when it can’t hit its KPIs organically, is “absolutely still common practice” in the industry, according to King.
“I find it fascinating that on LinkedIn I'll still see people advertising ‘I've got 500,000 hits a week, who wants to buy some traffic,’” he says. Although initiatives including ads.txt, supply path optimisation and MediaMath’s own Source have weeded out much of this.
The industry’s reluctance to uncover fraud in order to continue to trade on false metrics could be viewed as a type of fraud in itself, King suggests. He cites Facebook’s recent US$40 million settlement with advertisers for faulty video metrics that were reportedly inflated by up to 900% as an example.
“That's fraud, and there's zero repercussions for that. I haven't seen anybody reducing their Facebook spend or any attrition to Facebook’s share price,” he says, adding that the lack of independent verification on the walled gardens is a “huge issue” when it comes to the opacity of digital advertising.
Collaboration is key
Accountability, collaboration and education are key to progress, these industry experts believe.
“The finger is pointed at agencies because the role of an agency is to advise their clients on how to navigate the complex ecosystem, but the brands hold the budget that is being spent,” says Dewar. “A brand being better educated about where their money is being spent, whether in a positive or negative ecosystem, will help incentivise the rest of ecosystem to do a better job.”
Publishers also have a responsibility to put their own mitigation strategies in place, beyond simply dropping the ads.txt file on their web servers. We have already discussed in a previous feature the fact that ads.txt is an important but imperfect solution that reduces but doesn’t eliminate fraud from a publisher’s website. But publishers, who are often the poorest partners in the supply chain, generally treat fraud as checked off the list when they have employed ads.txt.
"It tends to be a game of whack-a-mole. It's a point solution for a very specific problem, and then you have very smart people doing very bad things on the internet, and they can figure out a way to get around that.”
—Zachary King, MediaMath
She also believes it is critical for the industry to agree on some common terms and solutions to help eliminate the complexity surrounding ad fraud.
“At the very least we need to come to some agreement on common terms and types of fraud to help educate the industry. It confuses everyone when there are different terms being thrown around that mean the same thing,” Quigley says. “How do we have industry stand up and say this is seriously a problem.”
King agrees the industry needs to get behind one standardised approach: “It’s very, very rarely black and white in this space, there’s a thousand different shades of grey with different parties proposing different solutions. Everybody has to sit around a table and say, ‘yes, we agree that this is the fraud-verification approach’ or you end up in that fragmented ecosystem again.”
This ‘stronger together’ mentality underpins MediaMath’s new digital media-buying framework, Source, which has brought together a range of advertisers, agencies, SSPs, publishers and verification partners to build what the company claims is a 100% verifiable supply chain. MediaMath has committed to running 100% of its campaigns on Source by the end of 2020.
King recognises that Source being a MediaMath product means it may not be adopted by rival DSPs and therefore may not be the only solution developed, similar to the way different companies and consortiums are attempting to develop the internet’s common ID. But he welcomes this potential.
“If it turns out that The Trade Desk or Xandr wants to do something similar then great, because it will mean we will get to that place where all of the big brands and publishers are operating in the same kind of way,” he says.
“Otherwise it tends to be a game of whack-a-mole. It's a point solution for a very specific problem. And then you have very smart people doing very bad things on the internet, and they can figure out a way to get around that,” he concludes.
If that's not enough to put you off measuring clicks, I don't know what is.
This piece rounds off a trio of ad fraud features. Check out the others: