It's been a constant theme of 2014, but I sense the discussion is now percolating beyond media and advertising types to those in other business functions wondering why on earth they would take a risk using programmatic and RTB in ad exchanges. The proof for me comes with The Times reporting research for FT.com that nearly three in four ad spots offered in open exchanges was fraudulent. Of course, the legitimacy of spots offered by The Rubicon Project, with which FT.com has a direct partnership, are not under question. It's those too-good-to-be-true $1 per thousand impressions (or CPM) that lurk on open exchanges that are clearly open to suspicion.
Thus, when FT.com had the open exchange inventory investigated, 72 percent was found to be fraudulent. California-based Pixalate tracked the spots to rogue "mask" sites that were pretending to be the FT.com to fraudulently attract bids from brands for its inventory that would almost certainly never be viewed by a human. The sites were found to originate in North Africa, East Europe, Asia and the Middle East. Not surprisingly, Brian O’Kelley, chief executive of AppNexus, told the newspaper: “We have to stop this now, for the sake of the internet.”
The comment is not melodramatic. It's a sign of how bad things have become in the automation of buying and selling digital display.
The elephant in the room here is that at the moment it is the advertiser that is paying the price. Agencies can carry on charging commission and ad exchanges can carry on delivery impressions, but it's the advertiser who picks up the tab for the fraud. Ironically, it's the advertiser who has turned to an agency so their interests can be looked after and they can carry on running their business, safe in the knowledge they are in the hands of experts.
There are many things that exchanges and agencies can do to protect against budget going to such unscrupulous sites. Programmatic systems must be set up to know which exchanges are the legitimate preferred supplier of each publisher's top inventory so that deals that are simply too good to be true are not offered. In fact, the technology could surely be programmed not to offer deals that are too good to be true, or at least flag them up as way below what the market would normally expect legitimate inventory on the suggested site to go for. Agencies and advertisers could then elect to avoid bidding on inventory that comes with any such warning.
Presumably, IP addresses -- although constantly changing -- could be blocked in real-time far more proactively by all exchanges and not just the top tier of best performers. The most obvious step would also be to instruct programmatic systems that there is no such thing as advertising on bbc.co.uk. Thus, any exchange that is allowing space on the advertising-free site should be avoided.
Ultimately, what needs to happen is that the money is followed and some crooks are put behind bars. At the moment, it appears to be just too easy. Experts warn that someone with limited knowledge can buy a rogue site and get listed on ad exchanges with very little effort and begin defrauding brands immediately. Compared to, say, launching a denial-of-service attack and demanding a ransom to allow a site to get back on its feet, programmatic fraud just seems far too easy. You don't need to know what you're doing, and you don't need to break cover to demand money. Instead, it just flows to you a click at a time with only the most scrupulous brand, agency or publisher noticing.
While predictions for 2015 might make great reading at this time of year, the digital display industry has a far more pressing task. If it doesn't clean up the technology so it only works with legitimate sites, one of two things will happen. Brands will insist their budget is spent direct with publishers or they will pull out of display.