One of the investment themes I’m currently working on is in the TV ad buying category. The TV ad buying space is littered with tales of companies that have failed to gain much traction. TV buying today looks very similar to TV buying 10 years ago. Why do we think that might be about to change?

There are a few reasons:

1.  The arrival of "TV everywhere" means that TV buyers and marketers are being forced to look at non-Nielsen viewing metrics to gauge overall campaign performance. That opens the door for marketers to consider set-top box data, social data and other sources as valid targeting and reporting mechanisms for traditional TV advertising.

2.  TV media buying margins have come under intense pressure from buyers and are now wafer thin. This cuts both ways: it means TV media buyers have no room to experiment, but it also means that the more forward-looking agencies are focused on finding new ways to bring value to their clients that can provide them with margin expansion opportunities.

3.  Online performance-based buying is now well-understood at the upper levels of media buying agencies. The barriers to applying similar models to TV due to cultural and expertise inertia are slowly wearing away.

Improved TV ad targeting can be divided between two broad approaches:

The first is addressable advertising that targets ads to individual users or households. This makes TV advertising much more effective, but can pose significant inventory management challenges on linear TV − if an advertiser targets pet owners with an ad, which ad does the rest of the TV population get shown?   The TV business is a long way from being able to handle this kind of inventory fragmentation effectively on linear TV. Perhaps there’s an opportunity to aggregate inventory across non-linear viewing on DVRs, VOD and second-screen devices.

The second approach is "audience targeting." Here, instead of an advertiser buying a show or a daypart, they specify the characteristics of the target audience they want to reach and the targeting solution leverages detailed set-top box viewing data (as well as other data sets) to identify particular ad slots in the schedule in which the target audience is likely to over-index. These over-indexing spots are then mapped against cost to design the most effective ad schedule. Because the advertiser is still buying the entire audience, some of it falling outside the target, this method is perhaps only 20 percent more effective. But, it’s more easily adoptable because it doesn’t disrupt any of the established buying processes. 

TV advertising has sustained so well because it's one of the most effective promotional formats we have − full sight, sound and motion with no distractions. For the future, it is clear that new data sources and new ways of interpreting and applying that data will be increasingly critical components supporting the value of the TV market.