• by Brian Wieser , March 28, 2014

This week we attended parts of two events related to our universe of coverage, including Adobe’s “Digital Marketing Summit” and AdExchanger’s “Programmatic IO”. In watching some of the presentations and in speaking with practitioners in and around these conferences, a common theme of discussion related to that bugbear of a topic, agency disintermediation.

The notion that media owners such as web publishers might sell media directly to the largest brand marketers on a widespread basis has always seemed improbable to us. The notion of shifting away the work of an agency has most probably been around for at least as long as the industry, er, intermediated itself between publishers of newspapers and marketers during the 19th century. Google’s rise over the first half of the last decade followed by their efforts to sell print and radio to advertisers directly and actions such as Microsoft’s purchase of the agency Razorfish (as part of its Aquantive acquisition) helped amplify latent concerns that agencies might not actually be producing enough value to justify their existence when compared with the power of modern technology companies.

People with these perspectives failed to recognize that technology companies often focus on technology and media companies tend to focus on media rather than agencies which tend to focus on client service first and foremost across virtually all agency disciplines. The primary service that agencies offered then and offer now is the socialization of ideas across complex marketer organizations, whether on the creative, media, CRM or other part of the business. The more complicated the marketer organization, we would suggest the more reliant it is on third parties to provide and socialize ideas.

The proliferation of programmatic trading and the rapid evolution of the agency trading desk seems to be bringing disintermediation concerns back to the forefront. Much as agencies salivated at the reported margins that ad networks generated before creating trading desks to serve as their own de facto ad networks to disintermediate some of their suppliers, some marketers have sought to disintermediate their disintermediators by establishing their own “brand trading desks” using technology which is increasingly available off-the-shelf or on a customized basis. The promise is amplified when we consider the potential to apply first party data in campaigns that might be too sensitive to share with an agency or other outsourced provider of related services.

Given all of this potential for cost savings and revenue expansion, large brands such as Procter & Gamble and Kellogg’s have been particularly prominent among the world’s biggest advertisers in establishing the capacity to buy programmatic media in-house. With such illustrations of successful brands taking the plunge in disintermediating their agencies, at least in part, interest in exploring the prospects of bringing programmatic trading in-house will inevitably expand.

But how durable is this approach and how widespread will it become? As we indicated above, the complexity larger brands must manage in the context of media fragmentation clearly requires expert integration, and few marketers will want to take on all of the responsibilities of their media or creative agencies, which means some trade-off between direct costs, effectiveness and marketer-specific customization. Best practices in working with certain media types and negotiating with certain media owners only emerge through scale that exceeds the size of the largest individual marketers. Further, access to talent and retention of talent is challenged when most of the tech and agency industry is concentrated in New York, San Francisco, Chicago and LA, but relatively few large marketers are based there. And finally, those appealing margins that attracted agencies to this opportunity in the first place? Such figures were usually gross margins, which were a long way from operating margins or cashflow. Maintaining a trading desk and relevant talent will be costly for most and may require more ongoing investment than a marketer is willing to sustain. Perhaps more critically, agencies serve as an external entity which can be varyingly used to validate choices a marketer wanted to make anyways and they also serve as a foil which can be blown out as needed in order to deflect blame when things go wrong. If a marketer drops the use of their outside agencies, they might lose their job sooner rather than later.

Some large marketers will overcome these issues and sustain internal buying of programmatic media, although it would seem unlikely to be widespread. It’s worth noting that the notion of in-house agency work is far from alien to the advertising industry. Many of the world’s largest marketers will maintain the capacity to buy newspapers in-house, for example. Web site design may or not occur within a marketer’s organization. And even media buying exists in-house for some large brands, most notably InBev’s Busch Media.

Walking around parts of the floor of Adobe’s conference this week, it was impossible not to note that this company – which owns a host of marketing technology businesses which could help to disintermediate agencies for marketers who want to do so – is embedded with agencies in many ways. Many marketers who want to go it alone have more and more tools to do things themselves. But these customers were probably not particularly critical for agencies anyways. The bigger impact of the presence of marketing technology companies is automation of processes will help drive down like-for-like costs for like-for-like services. Today’s agencies will mostly make up for that by providing increased numbers of services to (potentially) an increasing number of discrete advertisers. That that fragmentation exists and persists is what will keep agencies going strong so long as they continue to provide value in their roles as intermediators into the future.