Programmatic buying coming to TV

0

Coady DiemarEarlier this year, Interpublic Group announced that it was working to automate more of its media transactions. To this end, IPG announced that it would like to see 50% of all its media buys, not just online buys, transacted in a programmatic manner by 2016. According to IPG’s subsidiary Magna Global, there is a need to automate the buying of both national and local inventory in most media, including TV and radio inventory. We estimate the size of the media marketplace to be approximately $182 billion in 2012, including print media, consisting of newspapers, magazines and yellow pages (none of which are mentioned in IPG’s need for automation). For those not familiar with automated buying, one should think of machine to machine transactions, with computers making trading decisions within milliseconds.


During a trading desk panel at Advertising Week in New York City, executives from each of the agency trading desks were asked how much of each media dollar goes toward automated or programmatic buying today, and how much would go towards automated buying in 2015. Not surprisingly, IPG’s SVP of programmatic buying stuck to his company’s party line: 50% of buys would be automated in 2015, despite the fact that only 5%-7% of IPG’s media buys are conducted in a programmatic manner today.

Other agency trading desk executives were not quite so optimistic. For example, WPP noted that less than 5% of trades are conducted in a programmatic manner today, but projected that just 7%-10% would be automated in 2015. The WPP executive cited that, as the industry’s largest agency responsible for $90B in ad spend, “it would be impossible to automate half that in the next two years.” As shown in the chart below, other heads of agency trading desks believe that 40%-50% of all digital media buys would be automated in 2015, but not likely all media buys (Havas’ trading desk suggested it would be approximately 15% of all media buys). Regardless of where the industry finds itself in 2015, one thing that isn’t disputed is that automation is coming.

It is not surprising that IPG’s Magna Global is bullish on programmatic buying. Magna Global’s recent 5-year forecast showed programmatic buying as the second fastest “medium” over the next five years, growing at a compound annual growth rate (CAGR) of 25%. Only mobile advertising, projected to grow at a compound annual rate of 31%, is projected to grow faster than programmatic buying.

At a recent conference on programmatic buying, host company OpenX noted that the machine to machine buying and selling of advertising inventory is likely to have the same effect that machine to machine buying of stocks had on Wall Street exchanges, as trading moved from floor exchanges to electronic trading:

  1. The number of trades will likely increase
  2. The bid/ask spread will narrow (on Wall St., spreads that were as high as $0.25 a trade 20-30 years ago have fallen below $0.01 per trade today).
  3. Transaction costs will decrease; and
  4. The size of the market will expand

Certainly, agencies will welcome the lower transaction costs that programmatic will likely entail. Fewer people will be required to transmit paper insertion orders (IOs) once workflow automation has been brought to bear. Fewer sales people will be required as machines take over a greater percentage of trades. While sales forces may shrink, those in sales will focus on value added decision making, rather than processing orders.

Automating Online Display Advertising

The automation of $19 billion online display advertising is well underway (search makes up the remainder of the $36 billion online advertising marketplace). One subset of programmatic buying is realtime bidding (RTB), in which online buyers and sellers participate in auctions to clear inventory in realtime (the time it takes your browser to load a website). RTB has grown from 4% of all online display advertising buys in 2010, to 16% of online display ad buys in 2012. Recently, IDC released its latest view of real-time bidding and the research firm now projects real-time bidding to grow at a CAGR of 48% over the next 5 years, from $2.0 billion in 2012 to $14.4 billion in 2017. If this outlook holds up, real-time bidding as a percent of all online display ad buys would grow from 19% in 2013 to 41% in 2017.

Some project that RTB will come to TV as well, but given that most premium TV inventory is sold months in advance, if an RTB market for TV develops, it will likely take place in smaller, lower rated cable networks, rather than on the highest rated TV networks.

IPG Announces Programmatic Consortium

In another step towards an automated future, in August, IPG’s Magna Global announced the formation of the MAGNA Consortium, a group comprised of A+E Networks, Cablevision, Clear Channel Media & Entertainment, ESPN, AOL and Tribune. These six media companies announced that they will work with Magna Global to “establish an industry leading end-to-end, integrated programmatic buying business model to drive critical automation advances across all media transactions.” The group’s goal is to substantially improve the media planning and buying process, and will work to deliver:

  1. More precise targeting and greater performance through the greater use of data
  2. Streamline media negotiations through the use of data and technology-based buying platforms
  3. Apply technology solutions to automate legacy buying systems across traditional and new media

IPG expects that available inventory will include display, video, mobile, digital-out-of-home, radio and TV. The benefits to sellers are projected to be lower transaction costs, while the benefits to buyers are projected to be greater precision and efficiency in reaching their targeted audiences.

Programmatic TV Buying

The speed at which programmatic buying has increased in display ad buying has led many to forecast that programmatic buying is coming to TV as well. While programmatic ad buying is already coming to TV in the form of automating workflow, there are certain elements of programmatic buying that remain years away. For example, audience ad network Simulmedia sees four features of programmatic buying:

  1. The ability to incorporate granular data beyond what Nielsen data does today
  2. Media that is bought and sold using computer programs
  3. Dynamic ad insertion; and
  4. Real-time bidding

Currently, the first and second elements are already being incorporated into TV buying by companies like Simulmedia. However, TV currently is not designed for the third and fourth elements, at least not at scale. There is some level of dynamic ad insertion taking place as it pertains to video on demand (by companies like Black Arrow). However, dynamic ad insertion has not taken hold in live linear TV. Finally, real-time bidding of TV inventory may never come to pass, at least it pertains to premium television inventory. For example, it would make no sense for a large Hollywood studio to wait until Thursday night prime time to bid in real-time on available inventory ahead of a new movie release the following day. Chances are there wouldn’t be any available inventory, at least on premium TV networks.

More Metrics Coming to TV

In the same way that online buying moved from buying sites to buying audiences, TV buying is likely to move in the same way as more granular data (i.e., matching offline purchasing habits to viewing habits) is incorporated into TV buying. As shown in the chart below, there are significantly more metrics applied to online advertising and buying than there is in TV currently. Like display advertising, online video advertising is already being purchased in this manner. Meanwhile, TV is basically measured using reach, and sales are only indirectly correlated to offline sales, though this is beginning to change with the use of set-top box data. The common metrics used in display and TV advertising are shown in the chart below.

However, in the near future, significantly more data and metrics will be incorporated into TV buying, as shown below. Certainly on the agency side there is a real desire and a focus on bringing end-to-end measurement from the top to the bottom of the purchase funnel across TV. The hope is that as TV measurement becomes more granular and is married to purchase intent or purchase behavior, that both sides (online and TV) will begin working along common metrics which will foster more integrated campaigns and cross-platform ad buys.

IPG Partnering With Leading Tech Firms

While a full-blown programmatic buying process in TV at scale may be years away, the first steps are being taken, particularly in the area of workflow automation. IPG recently announced that it is working with Visible World and Adap.tv, recently acquired by AOL, to automate the TV buying process. Visible World, a targeted TV ad provider that enables advertisers, agencies, and media companies to deliver addressable, measureable ads, recently spun out a separate company called AudienceXpress to focus on the issue. AudienceXpress is a programmatic TV buying platform that automates the planning, audience buying, optimization and daily reporting for TV campaigns.

Other TV Software Firms Get in the Game

IPG and its partners aren’t the only ones focused on this issue. Long-time broadcast ad delivery company Harris Broadcast announced that it is partnering with placemedia to build a comprehensive advertising exchange for TV. With 97% of all video consumed on TV, placemedia has launched a fully-automated portal so that advertisers can plan, place, copy, report and buy and sell targeted TV ads.

Not to be outdone, TiVo, acquired TRA in 2012 and launched Tivo Research and Analytics. TiVo is combining Nielsen data to TRA data, which consists of matching TV exposures from 1.5 million homes with specific purchase transactions to bring internet-level measurement and accountability to the TV ad industry, while enhancing TiVo’s position in the billion dollar TV analytics business. To sum up, several TV-centric companies are focused on bringing better data and increased automation to the TV buying and selling process.

Online Video Companies Get Into the Action

Given their experience in bringing automated buying to online video, we expect several online video ad networks to look to bring programmatic buying to TV. It would not be surprising for companies like Videology, Collective Media, TubeMogul, Tremor Video or YuMe to look to bring programmatic buying to TV as the online video ad market is a $3.5 billion marketplace while the television ad market is a $70 billion marketplace.

It will be interesting to see whether companies that have had a traditional focus on TV (i.e., Harris, Visible World) or the online upstarts (i.e., Videology, Collective) will win the day. One would think that the online video companies, who have already solved many of these issues online would be well positioned to disrupt the TV marketplace. However, the TV world is a fairly closed marketplace, with fewer than a dozen companies that provide the majority of the networks that consumers watch.

The traditionally focused TV tech firms already have established relationships with the supply side and “speak the language” of TV. Meanwhile, online companies often talk a language that is unfamiliar to traditional TV networks. The question remains whether the TV-centric companies (Visible World, Harris, TRA) or the online centric companies (Videology) will become the underlying technology behind the increased automation of TV buying.

Changing the Way Ads Are Bought and Sold

Regardless of whether traditional TV tech firms or online ad firms solve the programmatic aspect of TV, the way in which TV inventory is bought and sold is likely to change dramatically over the next 10 years. For example, in the old days, TV inventory was bought via relationships, through person-to-person negotiations, and buyers bought programs as that content acted as a proxy for audiences.

As TV moves to programmatic buying, media buying will be bought and sold based on data and scientific relationships, conducted through machine-to-machine buying and selling, using real-time programmatic technology, and the focus will be on audiences, not programs as a proxy for audiences.

The Silo’d Approach Goes Away

As shown in the chart below, the TV and digital video marketplaces essentially operate as separate silos today. However, they are likely to come together over the next several years, as advertisers begin to buy across screens. Simulmedia estimates that virtually 100% of the $6 billion currently spent on online video ads is purchased using granular audience data. In the chart, this is demonstrated by the digital video circle being all blue (with blue representing the use of granular audience data as well as behavioral or intent data). Approximately 10% of digital video buys are conducted in a programmatic fashion (with white dots representing programmatic buying).

By 2016, Simulmedia projects that programmatic buying could account for three-quarters of these digital video buys (as shown by the white dots) at which time the digital video market will have grown to become a $16 billion market place. By 2020, as the online video sector grows to become a $33 billion marketplace, programmatic buying could account for 95% of digital video buys, and the walls that separate the TV and digital video marketplaces will have come down, with advertisers now buying across screens as they follow audiences from online (desktop/smartphones/tablets/connected TVs) to offline (TV).

The chart also shows that audience buying and programmatic buying will likely take significantly longer to play out in traditional TV. For example, in 2013, it is estimated that only a few hundred thousand dollars of TV buying is based on granular audience data (as shown by a very small blue dot in the lower right corner of the $69B TV market). Currently no TV buying is conducted programmaticly. However, by 2016, roughly one-quarter of the projected $75 billion TV market could be purchased using audience buying mechanisms (represented in blue), while perhaps $3 billion of the markets’ buys could be conducted programmaticly (represented by white dots). By 2020, as the TV market grows to $83 billion, a significant majority of the TV market could be purchased using audience buying strategies, while a little less than of all TV ad buys could be conducted in a programmatic manner.

Final Thoughts

Broadcast and cable TV operators have been watching the evolution of the online advertising industry, and they see that change is coming and that programmatic buying is likely to be the way of the future. At the same time, history has shown that while programmatic buying of online advertising has increased, prices for available online inventory have continued to decline. Does one cause the other, or have online prices declined as a result of the near unlimited supply of inventory online. The question is whether TV, whose inventory levels don’t change with the size of the audience, will witness some of the same commoditization that has taken place in the online world.

As a consequence of the long-term price declines witnessed in online inventory, we don’t see TV companies rushing to embrace programmatic capabilities. That is not to say that TV companies won’t embrace automated workflow improvements that make TV easier to buy. But creating automated exchanges where TV inventory can be “cleared” may take longer to play out, except perhaps for the lowest rated cable networks. In the end, the market will only move as fast as programmers allow the market to move. Stay tuned.

Coady Diemar continues to meet and work with some of the most innovative digital media companies in the sector. We welcome the opportunity to share our ideas and relationships to help companies turn their challenges into opportunities as well.

–Chris Ensley, investment banker at Coady Diemar Partners
Coady Diemar is a boutique investment banking firm providing mergers and acquisitions, private capital markets and strategic advisory services to growth companies in digital media, mobile and other industries.
COADY DIEMAR PARTNERS
1370 Avenue of the Americas
New York, NY 10019
(212) 901-2600
Reprinted with permission.