Commentary
Will Rise Of The Programmatic Machines Spell The End For Media Agencies?
- by Sean Hargrave , Staff Writer, July 2, 2015
When you consider that half of British display is now bought programmatically -- three-quarters forecast within two years -- then it's not hard to see why a brand marketer may feel that having a string of agencies is an expensive luxury. One the one hand, data hubs can be set up to know what demographic box a person falls in and what they're interested in, judging by their recent activity. On the other, this data enables target lists to be drawn up to whom apt display can be fired off with less need for smart buyers and planners to become involved.
But hang on a minute -- is anyone else hearing an alarm bell ring here as ad fraud figures keep on suggesting there is a huge problem in letting technology take over? Nobody can truly know for sure how big a problem it is, but click fraud appears to be a billion-dollar industry which some extreme estimates suggest can account for anything between 15% to 50% of display budgets. With variances so wide, it's hard to pinpoint a figure, so the only real conclusion is that it's a massive problem that will only get worse. Why worse? Well, I see very little evidence that it has been tackled globally, and as the rise of programmatic leads to a wave of new money coming into automated display buying, it's not hard to imagine that this will create more opportunity for fraudsters.
The trouble for media agencies here is that ultimately, they are pretty powerless. They are the middleman who oversee campaigns. However, the issue does raise a huge question mark over advertising networks and publishers selling their inventory through automated platforms. Will they both grow a spine and begin to investigate one another and their sources' traffic to make sure they are legitimate? Will they sign up in great numbers to schemes, such as that run by the IAB UK, to show they are taking these steps and so are more reliable than companies which do not?
The quandary for media agencies is that certainly in the case of the large holding companies, they now have massive trading desks that global giants are throwing huge display budgets at. Do they truly need to lift up any rocks looking for bugs? Or should they sit back and enjoy the double-digit annual growth that most are reporting? If they go fully legit and question everything, will they have enough inventory left for the swelling demand caused by the huge rise of programmatic?
At some point, agencies will have to be perceived as not only the clever guys who know how to interpret data and turn it into cost-effective programmatic campaigns, but as the police officers who can pretty much guarantee a brand the media they're buying is being seen by humans.
But therein lies another problem. If there is such a guarantee that media has been seen by humans, it is likely to come from a tool that measures behaviour (a "customer" who constantly logs onto multiple page for just a second is likely to be a robot) and only allow media to be bought via companies that are signed up to a code that compels them to vet traffic and publishers effectively. When this becomes achievable, it's not hard to imagine that it will be made available as a single tool that brands can simply use themselves. Many are already building data hubs, and some are buying their own media and learning how to deal with going direct to market. If they have a means to better qualify traffic sources and judge an ad network's efforts in routing out fraud, then that would be a significant enabling tool that could persuade more to go straight to market.
When technology is how display is bought and sold, and hopefully also the answer to cleaning up the process, it's hard to validate your existence on the human qualities or the experienced wise people you employ -- particularly when ad fraud has ballooned, often unchecked, on their watch.
Sean Hargrave is MediaPost's London Editor. You can reach him at seanhargrave@btinternet.com.
They'll have work to do certainly justifying their position, though I'm sure they'll find a way to adjust their models so they are not cut out one way or another.
The reaction from some agencies thus far, though not all, is not necessarily that sophisticated though. They simply demand full site/app-level transparency for brand clients, as if that magically solves all brand safety, viewability and fraud concerns. If questioned they make vague noises about pirates and various nasties lurking in dark corners and how they couldn't possibly allow their clients to go anywhere near it.
That is the equivalent of only eating at 5 star restaurants - and never at the local kebab shop God Forbid- in the belief that they are immune from food poisoning, when of course they are not. It may improve the odds but you'll pay through the nose, and now and then when you do want a kebab, the local guy might in fact have the better, healthier, cheaper product in that specific case.
They key point being neither "premium" or "transparent" automatically equal 'safe' nor 'quality'- depending on your definition of quality as far as your business objectives go.
The more important thing is to work with people who are *transparent* in terms of their tech and processes, know where your ad is going, and who are accountable for delivering you the product you pay for. With programmatic the risks here are greater and so need to be taken even more seriously, so the industry finally waking up and creating standards and means of verification is a good thing certainly.
It's not clear whether (some) agencies are always being that proactive in educating their clients or whether they are simply following through on brands' knee-jerk reactions, but if brands ask their agencies more awkward questions about the value they are getting and the accountability in the process, and agencies focus on real 'lower funnel' results for their clients, money will naturally drift away from fraudsters to genuine suppliers who have their houses in order and work with the IAB, JICWEBS etc etc.