I tried to follow the logic of Jack Marshall’s piece, entitled “The Consequences of the Gift Economy of Online Ad Sales.” It goes something like this. Small publishers can’t compete with large ones when it comes to entertaining media buyers, and as a consequence, innovation in the media buying sector is being stifled. Fear not, though, because robotic media buying is here to save us from the corrupt 25-year-olds who run the media business.
I have no interest in re-fighting the media graft wars. We all know that there are ad agencies that could stand to tighten up some of their protocols when it comes to accepting gifts from vendors. It’s a big leap of logic, however, to say that lunches at Nobu are holding back innovation. And one needs to cross an even wider logical chasm to believe that automated buying would alleviate these problems.
If the “Gift Economy” is problematic, let’s first look at what Marshall suggests should replace it.
We’re five years into an earnest attempt to automate media buying through ad exchanges and the technologies that support them, and network/exchange average CPMs are having trouble getting north of the $1.00 mark. Depending on which source you believe, network/exchange CPMs are anywhere between half and one-tenth of a publisher’s average CPMs. And network/exchange CPMs are not projected to rise out of the basement anytime soon, when compared to the CPMs garnered by direct sales forces. I challenge anyone to find an unbiased source that claims otherwise.
While it’s easy for those making the efficiency argument to get excited about CPMs a tenth of what they are today, the reality of the situation is that few professional content producers can continue to do what they do if you cut their average CPM in half, much less if you reduced it by 90%. There’s no evidence to suggest that if premium inventory were to be offered on the exchanges, that premium pricing would hold. Why? Because using algorithmic buying masks the advertiser’s intent from the publisher, making it impossible for the publisher to understand who is interested in its digital ad inventory and for what reasons. Thus, pricing becomes a quick race to the bottom.
So, if you were to wave a magic wand and make digital media buying agencies and direct sales forces disappear, we would all get to witness the wholesale destruction of professional content production on the Internet.
Truth be told, the digital ad economy cannot exist solely on Real-Time Bidding platforms. Robotic media buying isn’t ready for such a shift in dollars, nor will it be in the foreseeable future.
Next, let’s examine the ridiculous notion that the algorithmic buying space is somehow a paragon of virtue when compared to human-engineered media buys.
My team and I have vetted hundreds, if not thousands, of ad products from companies that operate within the RTB landscape. The vast majority of them will not permit a peek inside the black box that constitutes their pricing model. Many of them don’t want to disclose how much they’re paying for media and how that compares to what they would charge my clients. Anyone who takes their duty to their clients seriously understands how appalling that is.
Regrettably, some don’t take it very seriously. When agencies decide to get involved in the automated buying business via an agency trading desk, serious conflicts of interest can emerge. Don’t take my word for it – the ANA advises its members to look into these potential conflicts and understand them fully before committing to this method of buying media.
While this pricing opacity is the norm, it’s ridiculous to make the argument that automated buying is somehow more virtuous than human-centric digital media buying.
It also occurred to me that while automated buying promises increased efficiency, the evolution of that business has been anything but.
When I saw my first LUMAscape, I thought it was a joke. That is, I saw it presented at a conference and started laughing. I saw a great irony in that first LUMAscape – the notion that something that was supposed to bring simplicity and efficiency to a marketplace was itself so complex and inefficient. Irony is not lost on me.
Terry Kawaja, the guy who invented the LUMAscape, told the crowd that day that consolidation was both necessary and expected in order for automated buying to work. Instead, with every revision of the LUMAscape, the slide grows more and more complex. The idea that the process of buying digital media is being streamlined is laughable if you look at the evidence that is on the table today.
I’d venture to say that much of the interest in algorithmic buying is driven by the influx of capital into that sector. You’d have to go back to the first dot com boom to find a comparable precedent. Ultimately, those who play in the space will need to decide whether they believe the money being poured into it is about streamlining media buying or whether it’s about capturing a greater margin for the companies involved.
I do not mean to suggest that algorithmic buying has no place in the business (it does). Nor do I want to suggest that it’s okay for agencies to allow their judgment to be influenced by vendor gifts (it’s not). But I don’t believe for a second that automation will save us anytime soon.