If there's so much ad fraud, where does all that money go?
I was recently asked a simple question, by someone relatively new to adtech -- "if there's so much fraud, who's making all that money?" My answer surprised him. My answer was "everyone." He was expecting me to tell him about crime syndicates, nation states, and master hackers like the ones aggrandized in TV shows and movies. But the proceeds of ad fraud are far more mundane and widespread than that. Any fraud investigator will tell you that the most reliable way to find fraud is to "follow the money." So let's do just that.
Hold-co media buying agencies
The largest of advertisers all use large media buying agencies to spend their digital budgets. They make a media plan at the beginning of the year and every day they watch the pacing like a hawk. Obviously, they should be watching the performance like a hawk, but instead the agency's job is to "spend it all." So they have to make sure the campaign "delivers in full." If they are falling behind on pacing, the agency gets into increasingly panicky situations where they have to take corrective action without telling the client -- like removing block lists, removing frequency caps, removing CPM ceilings, etc. All of these actions enable more impressions to be purchased so the campaign can "deliver in full" even if that means mixing in fraudulent inventory from fraud sites and apps. In some cases, the agencies desperately reach out to smaller or shady networks to get inventory for the end of the quarter or year so they can "spend in full." They explicitly tell the networks not to tell them where the inventory comes from; and those ad networks magically conjure up just enough ad inventory to eat up those budgets.
Don't tell me you haven't witnessed the above phenomenon regularly over the last decade.
Media agencies make money from ad fraud, even if they are not the ones deploying the bots themselves (hint: they don't have the engineering talent to pull that off themselves).
Publishers
What if you were a publisher, specifically a team member on the ad sales team. You were given a stretch goal of increasing ad revenue by X%. Your first thought is that the goal was insane and impossible, knowing that historically human audiences don't double in a year and their time-on-site or pages-per-visit don't triple. But there's a "performance bonus" tied to hitting those sales goals large enough that you can finally buy that Mercedes you've always wanted. What's the harm in having a junior ad ops person write some bots to repeatedly load webpages, or having content recommendation widgets added to the page to deliver guaranteed clicks, or perhaps just outsourcing the whole thing to a "monetization partner" who will take care of all those details and give you plausible deniability?
What if you weren't even a mainstream publisher with any scruples. What if you were an opportunistic entrepreneur who saw how simple it was to set up hundreds of thousands of fake websites using Wordpress templates, populate the sites with 100% plagiarized content, and monetize by selling ads through programmatic exchanges?
Don't tell me you haven't witnessed the above phenomenon regularly over the last decade.
Publishers, good and bad, make money from ad fraud, even if they are not the ones deploying the bots themselves. Some do. We call those "vertically integrated bad guys" because why pay an outside vendor for bot traffic when you can just as easily manufacture the fake traffic yourself.
Venture-funded ad tech companies
Ever wonder how all those ad tech companies can all show hockey-stick revenue growth and user growth? Human audiences and human customers don't move that way. Of course, we've seen a few examples over the years where lots of humans get enamored with a site (e.g. Groupon) or an app (e.g. Pokemon Go) for a period of time. But that interest wanes eventually and audiences move on. But for the most part human audiences change slowly and incrementally. Just ask any real, mainstream publisher about their audiences. If anything, they have seen audiences shift away to other channels and forms of media over the last decade.
So how do venture-funded ad exchanges show relentless growth? How do data sellers show large increases in user profiles they can sell for money? Why wait for slow humans to do the action you want them to -- e.g. visit your site or use your app? Why not just pay someone to acquire new audiences for you (audience extension) or acquire new app installs for you? Some vendors appear to show large numbers at low prices. It's almost magical.
An entire ecosystem of cooperative partners has been built up over the last decade. Every vendor is befitting from ad fraud directly or indirectly. The super large numbers of ad impressions, visits to sites, users of new mobile apps, etc. are all driven by bot activity and other forms of fraud. This is how company after company can show hockey-stick like growth patterns, as they promised in their investor presentations. The real world does not work that way. But in digital, it's easy to manufacture whatever digital metrics you or you investor wants to see. Of course the people who work at those companies don't know this is happening and are often not the ones doing the fraud. They just want to keep their jobs at the high growth venture-funded companies. Even if they see something shady going on, it's far more likely they just ignore it and look the other way than report it and risk their own job over it.
Fraud detection tech vendors
Even fraud detection vendors love ad fraud and benefit from it. After all they would not exist if there were no ad fraud. And they won't exist if ad fraud were ever actually solved. So while these fraud vendors are not necessarily committing the fraud themselves, they are certainly making revenue because of it. Both DoubleVerify and Integral Ad Science showed $100 million in revenues each, last quarter, as public companies. These companies rely on fraud to continue so they can continue to make money, even if their tech doesn't work that well.
So What?
I won't belabor the point any further. I assume you get that "everyone" makes money from ad fraud, even if they were not the ones making the fake sites and fake traffic. Even I make money because of ad fraud -- I advise clients that there IS ad fraud, that much of it is not caught by existing fraud detection vendors, and how to avoid the fraud by changing their media buying or blocking bad sites and apps from those buys.
A corollary answer that my friend was not expecting is that "we can't solve ad fraud." There's too many parties with too many vested interests (i.e. revenue) that want it to continue and need it to continue.
The real question is whether YOU want to do real digital marketing and drive real outcomes with the budgets you invest. If the answer is yes, we don't have to solve ad fraud for the entire industry or for anyone else. You can do that yourself, in your own campaigns, by upgrading your tools. The previous tools you were using are inadequate and don't reveal the extent of the ad fraud impacting your campaigns. I can show you what FouAnalytics can see, both in your ad impressions as well as on your landing pages. "If you can see better, you can do better." That's how we can solve fraud -- for one advertiser at a time -- because the industry won't solve it and none of the beneficiaries of it will solve it either.
Published by: Augustine Fou FouAnalytics - “see fou yourself” bought traffic = bot traffic