• by Brian Wieser , August 23, 2013

As the summer winds down, it would seem appropriate to reflect on a key trend of the season. In the music world, the psuedo-robot duo Daft Punk seemed to dominate, much as the psuedo-robot concept of programmatic media buying is probably the closest term resembling a "word of the summer" for the advertising industry. Several prominent news items this week alone touched on the topic, including NewsCorp announcing that it would launch the "News Corp Global Exchange" using supply-side platform The Rubicon Project to support the initiative. Many other publishers also believe that by discontinuing arrangements with third party ad networks and by taking more control over the inventory they sell programmatically (for example, by choosing whom they will sell to and by setting price floors), they will alter the deflationary conditions which are so pervasive on the web. However, our view is that they will need to "Get Lucky" for these trends to reverse.

While we think that may publishers with high quality content will have no difficulty entering into arrangements with agencies and their trading desks to trade programmatically and bypass third parties. They'll probably be able to secure minimum budget commitments and retain premium pricing, at least initially. However, in our view, certain trends are probably too late to reverse, as they have been building over many years. First, and in general, brand marketers are becoming more focused on buying audiences rather than contextual adjacencies. In other words, they are increasingly indifferent as to what content their advertisements appear next to (within certain thresholds for brand-safety, of course) and this means that advertisers have a better ability to walk away from any negotiation related to web-based inventory. Second, the shift towards programmatic buying and real-time bidding for web inventory means advertisers are cherry-picking optimal individual impressions faster than ever before. As marketers buy ad inventory one impression at a time (rather than in packages, as most conventional ad sales on and offline are made) they are able to identify whether or not they are finding their target audiences across multiple publishers faster than would have otherwise been possible before programmatic buying came along. Third, the use of retargeting as a tactic means that marketers are increasingly focusing on audiences that they are better able to track than publishers often can. A related consideration is that marketers and their agencies are increasingly using data management platforms which they control to better determine where they allocate their spending.

As a consequence of these trends, any commitments that publishers are able to secure in the near-term will probably fall over time unless pricing falls, too, or unless performance towards a marketer's goals are sufficiently strong that a site's context actually does matter. Among web publishers, News Corp probably has a better chance than most, especially for its former Dow Jones properties, which remain must-have buys for advertisers in certain categories. They also retain a significant volume of user data related to their subscribers' logins, and can undoubtedly provide top-tier targeting and strong performance.

But many other publishers are in a tougher spot: human sales forces are expensive to maintain, and can't necessarily turn around the aforementioned buyer preferences around audiences and retargeting which are becoming increasingly detrimental. And yet, programmatic sales aren't necessarily a solution which will lead to long-term growth. The options available to publishers are generally poor. Perhaps the best solution we have learned of is the notion of eliminating the sale of standardized display ads, and instead re-focusing on selling partnerships, brand integrations and other high-touch advertising products. This probably means a smaller business for publishers which embrace such a model, but it is also one which is probably more sustainable going forward, not least because the advertisers who will buy in this form are probably focused on the often intangible value of publisher-brand interactions (or the old "adjacency" model).

Many publishers will not have content which is sufficiently high quality to embrace this model, and others will not believe that their quality content is being commoditized as much as we believe it is. Will anything cause this trend to change? Unlikely, unless publishers can "get lucky" and change negative industry dynamics over a period of time that will last longer than a summer, and for many years more.

It's Totally Automatic (Not Totally Programmatic)

Partially related to the programmatic theme, Interpublic's Magna Global and media owners including A&E Networks, AOL, Clear Channel and Tribune announced the "MAGNA Consortium" this week. The focus of the announcement was to highlight joint efforts to better apply data to media buying choices, streamlining negotiations and working towards making existing buying systems more automatic.

Automation is the key word in the announcement here. Workflows required to negotiate and execute buys still result in substantial absolute costs, even if they are often low as a percentage of total media spending. Savings can accrue to agencies who can boost their margins, and/or to the larger marketers who will generally try to claw back efficiency savings when they can. Further, new workflows could better enable buyers to make smarter choices and better satisfy client needs. In context of IPG's media agencies, which have focused on incorporating performance-based pricing into their contracts while also drawing attention to their cost-sensitive clients that they are focused on efficient processes, such an initiative could be helpful with future pitches or client retention activities. All of this relates to an important, if often little understood, aspect of what media agencies do day-in and day-out with their operations.

However, we think the story made a stir with the press and investors (and us too) because of a conflation between this news and other initiatives related to programmatic buying. The distinction between Magna's efforts to automate and any initiative to drive programmatic buying into traditional media is significant, not least because programmatic buying of traditional media is unlikely to occur on a widespread basis for an extended period of time (at least in the United States).

Similar initiatives in the past made an effort to drive automation into TV and radio ad sales generally foundered, including those spearheaded by eBay, Google, Navic (founded by now-CEO of Aereo, but later purchased by Microsoft), dMarc (acquired by Google), SWMX (whose former CEO, incidentally, is now a senior executive at Rubicon Project) and Bid4Spots. None of these efforts were what might be called "programmatic", not least as it was not possible to purchase individual impressions. Instead, individual units might be isolated, but all consumers viewing an ad unit would be exposed to that ad, regardless of the advertiser's intention to reach that specific consumer. All of these efforts were eventually shelved, largely because of limited access to the most highly desired inventory. That the initiatives were not whole-heartedly embraced by media agencies at the time probably also made a difference.

But this news is somewhat different: it is an agency leading the initiative. To the degree the company is sufficiently committed and can convince its clients to be supportive, Magna's goals - which are to "deliver 50% automation to its buying process" by 2016 - are realistic, if still dependent upon finding increasing numbers of partners to work with towards its objectives. Accomplishing this will probably require significant human efforts in a process of persuasion which is unlikely to be automated itself.

This was excerpted from Brian Wieser's Madison & Wall report.