By Alexander Cherry - March 25th, 2015
What best characterizes media transactions in 2015: fragmentation or convergence?
What best characterizes media transactions in 2015: fragmentation or convergence? Answering “fragmentation” would certainly find a great many advocates. After all, the written word – like what you’re reading at this very minute – can be accessed on paper, on your desktop internet browser or within a mobile app, like LinkedIn or Facebook... Similarly, television is no longer just on your TV, but also on your desktop, your tablet and your smartphone. It’s not only live but time-delayed and DVR’d. Plenty of television isn’t even made by television producers any more. Similarly, the choice of media on offer now, whether print-style, video or audio, is staggering compared to even just 10 years ago. For an advertiser to complete the audience jigsaw out of so many disparate pieces looks a close to impossible task— so it seems logical to conclude that media transactions are becoming more fragmented. Despite sometimes appearing to add to the complexities, programmatic has actually emerged not as part of the problem but as part of the solution. The first DSPs bought display inventory across a multitude of desktop sites within a single platform, a task which would have been prohibitively time-consuming by direct means. Programmatic thus served as a “front end” for the fragmented supply sources of desktop display. Digital channels and screens share an infrastructure in a way that offline channels like newspapers, cinemas and billboards simply do not, so in principle what works for one digital channel/screen combo should ultimately work for another. So it is that, less than a decade since the launch of the first ad exchange for remnant display, programmatic spans the entire digital media ocean: whether it’s desktop video or mobile radio, impressions can be bought and sold programmatically on thousands of separate sites through a single central interface. So perhaps we should recognize that media transactions are actually becoming easier, more converged. That’s not to say there are no hurdles to overcome— far from it. Programmatic markets vary by content type; the desktop screen is notoriously difficult to reconcile with the cookie-less mobile screen; and we should not forget about the traditional offline world either, which in 2015 will still be worth a whopping $131 billion in the US, more than twice as much as all forms of digital put together (eMarketer, June 2014). Let’s cast a cursory glance over these new frontiers of the Programmatic Empire as it expands... Video as the New Display Programmatic was originally conceived for remnant display, a format that has always been characterized by oversupply due to the relative ease with which publishers can create new inventory. This problem has not historically existed for video and audio, due to the barriers to entry being higher and inventory highly sought after, which has resulted in slower programmatic adoption in these formats. However, applying programmatic rich-data insights to already high-impact formats (like video) delivers proportionally more overall impact than applying them to low-impact formats, like banners. Recognition of this fact is driving record demand for targeted digital video among brands. As a result of this, TV networks are seeking to monetize their OTT content through dynamic ad insertion, while many print-first publishers are scrabbling for high-quality short-form video. In the future we can expect video to no longer to be a novelty but rather a staple of users’ online content diet. And the greater the scale advertisers can achieve with targeted digital video, the more pressure there will be on broadcast networks to introduce programmatic-type solutions. Cross-Device Tracking A particularly well-documented challenge for digital marketers has been the leap from the desktop to the cookie-less mobile, and the difficulties tracking consumers on their journey between the two. The promise is sky-high because if marketers can follow consumers all the way along the funnel they can deliver them not just the right message but the right message in context. This would be a shot in the arm for brand storytelling and throw open the playing field for creativity and engagement. However, if advertisers’ customer-tracking models are cookie-based, as they often have been, then buying mobile is essentially buying blind. In the vendor community, competition to own this nascent mobile and cross-device tracking space is heating up, with a melting pot of probabilistic and deterministic solutions in the offing. Probabilistic device association, as the name suggests, establishes a high likelihood that two devices belong to the same user based on a range of non-cookie data points (such as time, location and wifi network). Deterministic device association does away with this uncertainty by relying on unique user IDs, but it only functions within a given logged-in ecosystem (such as the Facebook or Amazon online ecosystem). While deterministic association certainly unlocks considerable potential for cross-device marketing, its role in the long-term development of online media as a whole is perceived by many as ambivalent: for it looks like the crown of audience identification, once fought over by agencies and publishers, may ultimately go to the tech giants (who are simultaneously media giants). And these giants may rule their media kingdoms as walled gardens, with mobile identification possible only within, but not across, each ecosystem. This could limit the overall promise of cross-device tracking for marketers. It certainly might diminish agencies’ value proposition as a central hub for complex, multi-partner buys, and it might also force independent publishers to become tributaries of big tech as a means of enabling mobile and cross-device targeting on their sites. This consolidation could yet be counterbalanced by the development of a reliable and agnostic probabilistic standard, so this will certainly be an exciting space to watch unfold over the next 18 months. But what is sure is that the long-sought answer to the question of cross-device tracking will raise further, and still thornier, questions… Programmatic Offline Given the amount of prominence attained recently by innovations in digital ad tech you might think that offline media had gone the way of the dinosaurs. But TV ad spend in the US will account for $71 billion in 2015, print for $31 billion, radio for $16 billion and outdoor for $7 billion (eMarketer, June 2014)— even if the long-term trend is for them to diminish, these are hardly numbers we can ignore. So what does programmatic mean in an offline world? Offline media is not addressable in the same way as digital, so you are reduced to buying content with its total audience rather than individual impressions. And because all that can be known about the audience is known in advance, there is practically little place for RTB. Programmatic in this context relates firstly to data and secondly to workflow. For example, data can be overlaid on a network of billboards, a roster of TV commercials or a swathe of local radio stations and used to aggregate audiences defined to varying degrees of complexity (admittedly with some inevitable wastage); algorithms can optimize competing campaign proposals; and a central software interface can automate all booking, running and reporting. While this incarnation of programmatic is less glamorous than its digital cousin and still fairly experimental, it will definitely grow as a new generation of marketers come to expect standard processes and interfaces across all their media buys. Early adopters may steal an edge on the competition: only last month Time Inc. announced as an industry first that it would sell portions of its print inventory programmatically. And of course, any solution that can marry up offline data, however loosely, with online data will be welcome to marketers, for whom, at least in digital, a total view of the consumer is starting to become a possibility. So will we reach programmatic everything and, if so, what will it look like? To answer this question we should remember that what is revolutionary about programmatic is not so much its automation of human work but rather the non-human work it makes possible in the first place, namely granular, data-driven decision-making. And it has never made sense for this to be confined solely to remnant banners. So the race is on to cascade the benefits across the whole digital buy, both in established areas like display and in growth markets like video and mobile. Offline still largely holds its own, for now, but as the balance shifts between online and offline worlds, some form of reconciliation will be inevitable. This is the true programmatic disruption: it is not necessarily about everything becoming automated or RTB, it is about different channels becoming cross-compatible in terms of targeting data. According to this theory, the future of media transactions will see a data layer marked by increasing convergence atop an execution layer marked by increasing fragmentation. Luckily for us humans, the execution side will be taken care of by machines. The upcoming Incite Programmatic Summit will feature an hour-long panel on Programmatic Everything, covering everything from TV to radio, mobile video to print. A team of programmatic experts – Melissa Grady, VP Digital Acquisition @ MetLife, Tom Hagopian, VP Advanced Advertising @ DirecTV, Megan Pagliuca, VP & GM of Digital Media @ Merkle, and Arun Kumar, Global President @ Cadreon – will share their insights and experiences on moving beyond digital display and embracing the full potential of programmatic to one day cover media buys across all channels. Check out the full agenda to see what else you’ll learn.