The Wall Street Journal gets apocalyptic about online ad fraud, but gets the story wrong

By James Robinson , written on March 24, 2014

From The News Desk

This morning, the Wall Street Journal published an alarming piece of ad tech reporting. According to Internet Advertising Bureau estimates, the article says, 36 percent of all online ad traffic is fake. With the article very willing to remind us that the market for online ads is expected to cross $50 billion this year, the clear insinuation - which isn’t stated explicitly, but is never disputed - is that online advertisers are being collectively defrauded of $18 billion.

The point of the article is solid. Online ad fraud is an expensive issue damaging the reputation of the entire space. Advertisers are paying small fortunes for bots to click on their ads. The lack of transparency and reliability is dissuading new advertisers from jumping in. Click fraud, bots clicking ads and ads placed out-of-site on web pages compounds into a prickly problem. For digital advertising newbies, a few bad apples poisons the perception of the whole industry.

But even still, the Wall Street Journal’s numbers are drastically overblown.

Running the article by ad industry insiders, the consensus remark was that performance based advertising renders its concerns moot. If an advertiser pays for a campaign by each sale or lead it generates, it doesn’t matter if a million bots click through. It’s not relevant traffic. “Bots don’t check out,” one agency manager remarked.

According to the IAB, two-thirds of all online ads are priced this way. CPM pricing - essentially paying for every thousand ad impressions - accounts for most of that other third. So, in one strike, the tainted piece of the pie becomes roughly $17 billion, not $50 billion.

Examining the problem outside of pricing models, the Journal gets it wrong again. The two largest players in the market, Facebook and Google, have fraud ratios that are vastly better than its scaremongering figures. Google accounts for a third of the online advertising market. It estimates click fraud to be less than 10 percent of AdWords traffic and doesn’t charge clients for the fake clicks it picks up on. Facebook estimated this year in its annual financial filing that between 5.5 and 11 percent of its accounts were fake and has shifted aggressively toward performance-based ad offerings.

The Journal’s own article walks back its extreme claim. It has no data passed the IAB’s claims to corroborate the 36 percent figure. It cites research from White Ops that puts a $6 billion price tag on online ad fraud in 2013. Which is damning of its own fruition, but by the Journal’s reckoning puts fraud at about a third as rampant as its own headline suggests. Display ads sold online through third-party exchanges are the most vulnerable to fake traffic, but the Journal’s own data embedded into its article puts this at 28 percent of the $50 billion pie.

Ad fraud is a problem and companies like - recently purchased by Google - and WhiteOps will always be running to stay a step ahead. But its an issue of evolution. Facebook used to be based around fan acquisition and companies building up platforms there by paying for likes. That system was easy to game, but the company moved way past that to offer up campaigns where brands were not just getting eyeballs, but demonstrations of actual results. Likewise, it’s a good bet that brands paying good money for display ads in weird parts of the Internet is also headed for the ad tech scrapheap.

[illustration by Brad Jonas for Pando]