The more traffic a site gets, the more money it makes. And sites need to get lots of traffic—lots more traffic than traditional media used to get—to pay for the costs of producing content.
This is more or less the justification for the somewhat scammy things that blogs do to acquire millions of extra page views every month. It’s why they publish first and edit or fact-check later. It’s why they admit that truth is irrelevant to an article’s value. It’s why they fall over themselves to cover pseudo events and marketing fluff. It’s why they split articles up into slide shows or paginate them. And of course, advertisers go along with this, because it means more opportunities to plop cheap ads in front of trapped users.
All these things are clearly manipulative, but at least they are not illegal or outright fraudulent. But what if prominent advertising networks and media sites decided that that wasn’t enough? What if they’ve decided to open it up to plain deception to wring extra money from brands?
As it turns out, this is exactly what is happening. A recent study by comScore found that 54 percent of display ads shown in thousands of campaigns between May 2012 and February 2013 never appeared in front of a human being. Rather, the traffic came from bots. As an advertiser, this would be like buying a billboard you were told was seen by thousands of cars a day only to find out that was because the billboard sat next to the assembly line at a Ford plant. Sure, that’s a lot of cars, but there’s no one in them.
In fact, the system is so broken that, for some publishers, knowingly buying traffic that comes from bots is part of their business model. An anonymous publishing executive, who claimed to be buying up to $35,000 worth of traffic per day, recently told Digiday that for publishers running an arbitrage model, all that matters is profit; quality of traffic does not factor into the equation.
The game is thus: You can buy traffic for less than a penny per visit from a vendor of dubious quality, because you know you have essentially an unlimited standing offer from advertisers that value those impressions at a penny plus. Of course, they have no idea they’re being hustled. They have no idea that a significant amount of those views never actually happened, despite what their servers log.
But what exactly is this bot traffic? Traffic from bots is the segment of Web traffic that is generated by automated botnets. A botnet is fake traffic that is generated from thousands of infected computers, usually via malware, aimed at generating fraudulent ad revenue through fake clicks and impressions.
Botnets usually originate in faraway countries like Estonia, Singapore and China (where labor is cheap). The largest cybercrime takedown by the federal government involved a group from Estonia, who used a botnet to make $14 million, with fake ads that appeared on ESPN.com and Amazon.com.
When a publisher knowingly or unknowingly buys botnet traffic from ad networks, what it gets in return is dirt-cheap traffic, usually for a penny or less. If questioned, an ad network would describe the traffic source as of an “unknown quality.”
It’s a classic case of getting what you pay for. For traffic that cheap, no publisher is so naive as to think that all of the traffic it just purchased is legitimate. But when you’re playing the ad arbitrage game, those margins do look good for the bosses in the C suite.
I’ll give you a very simple example of this type of transaction working on a much smaller level. When I publish this article, I have two options to help get it rolling with audiences. I can share it with my fans and submit it to sites like StumbleUpon, or I could go on Fiverr.com and buy fake likes, tweets and traffic to make it instantly one of the most popular articles of the day on The Observer, which will then earn the article more visibility and organic, real traffic.
Of course, there are other sorts of “fake” traffic too. I can buy clicks cheaply from Outbrain or Taboola. Even Google sells views for YouTube videos via Trueview. An ad agency or publisher can tell its client that a video or a branded piece of content did however many thousands of views when really those eyeballs were purchased at a $.01 per view from viewers unknown. It’s more legit than outright bot traffic, but we all know it’s not quite real.
A similar scam is being run across the Web by bloggers, publishers and marketers all the time. In my hypothetical, my incentive was attention for my article—and as a writer, I’m not super incentivized to deceive myself (though, if I was paid by the page view like some bloggers, I might want to trick my bosses). But for publishers and ad networks, buying cheap, fake traffic is an easy way to make a profit in an environment with increasingly low margins.
Typically, it is shoddy content publishers that utilize these dubious sources of traffic, but major media publishers have also been suspected of buying fake traffic. For example, men’s sites like Break, Crackle, Complex and Entrepreneur were found to have a high percentage of visitors who also visit sites that profit from bot traffic. There was also a recent controversy on YouTube when popular music videos suddenly lost a cumulative 2 billion views from their view counts—a result of Google removing suspected bot traffic.
For some ad exchanges, dealing in suspect inventory can mean millions more dollars at acquisition time. Earlier this year, AOL agreed to acquire Adap.tv, a video ad exchange, for $405 million. Some experts have said the amount of suspect inventory bought and sold through Adap.tv’s exchange could be anywhere from 30 percent to 80 percent. If true, that’s about as clear a case of the tail wagging the dog as you can get.
So what is to be done about all this supposed fraud? Some of the largest players with enormous ad spends have been going back to publishers and ad networks demanding some kind of “viewable guarantee.”
But the vast majority of ad buyers who can’t force guarantees may have to start picking their poison. Do they keep buying millions of display ads, knowing they’re lighting half of their money on fire? Or do they give that money to Google and Facebook,whose near monopolies on search and social mean they can demand increasingly higher rates? Some say that if online ad fraud persists, Facebook and Google will further cement their positions, at the expense of other cheaper ad networks.
I buy a fair amount of online media for my clients, and I have passed on certain vendors after seeing their names pop up on various lists of bot-traffic offenders. At the same time, what client wants to hear from their ad buyer that rates are creeping up because they’ve decided to be conservative with their negotiations?
The online ad game is corruptive, because the parties involved have little incentive to be honest. Ad networks have little incentive to clean up their exchanges, because they get a cut of every transaction, and cleaning up their traffic sources would lower revenues. Publishers don’t want to see penny-per-click prices disappear, because then their revenues would plummet and, more importantly, so would their traffic stats. Ad buyers know that their clients want the cheapest rates possible and the largest reach. At the end of it, it’s brands and businesses that get scammed, wasting millions of dollars on bogus, pointless advertising.
Like so many of the problems in online media, this kind of fraud is insidious, endemic and mostly ignored by the public. Everyone wants to celebrate or gawk at the monthly numbers posted by new media companies. No one wants to get into the nitty gritty of how they happen to achieve those numbers.
What kind of journalism, news or entertainment do you think a system like this is capable of producing? To me, it’s all fruit from the poisoned tree: bad incentives, a bad business model and a bad output.
Ryan Holiday is a best-selling author and adviser to many brands and writers. His newest book, Growth Hacker Marketing: A Primer on the Future of PR, Marketing and Advertising, focuses on the untraditional tactics behind a new class of thinkers who disrupted the marketing industry.