Movies and TV programs have featured their share of robots over the years, some of them kind and some downright creepy. The robots from the Star Wars’ clan are some of the good guys: R2D2 is the ultimate buddy and co-pilot, and Lieutenant Commander Data, from Star Trek the Next Generation, is a computational wiz and overall great COO. Rather than being menacing, these bots are always ready to lend a hand.
If we lived in a world with robots like R2D2 and Data, we could call on them to help us out with a lot of things - including entertainment. We could rely on their robot intelligence to intuitively know our entertainment tastes and preferences and calculate just what we wanted to watch. Data would have the knowledge to tailor viewing to meet our every mood, at any given time of the day. R2 would use his holographic projector to deliver perfect, personalized video content-- anywhere. And he’d be your personal gatekeeper, never allowing shows you didn’t like; instead he’d beep in “these aren’t the shows you’re looking for.”
While we may be light years from having personal robots, we are getting closer to one-to-one personalized TV and video. Today, most major broadcasters from Netflix, to Comcast, to YouTube and beyond, are using digital technologies to create personalized viewing experiences and help consumers search, manage and discover programming. As TV and video viewing options continue to multiply like rabbits, and technology advances, what will personal TV look like?
To explore new worlds...
TV and video personalization -- search, recommendation and discovery-- were born of necessity. Today’s digital video content and TV programming is much more prolific, more diverse and widespread than even a decade ago. What began as one screen and three networks now spans hundreds of subscription cable and satellite channels, thousands of premium and on-demand selections, dozens of pay-per-view offerings, and hundreds of connected TV apps.
Nearly all that TV content can also be streamed to a variety of digital devices. Add in streaming video providers’ programming (Netflix, Hulu, YouTube), online-only TV channels (Veetle and similar) and one-off programs, and the number and variety of choices makes selecting a program akin to looking through a telescope -- the viewing options seem unlimited. Viewers need discovery to manage, enjoy and explore programming; broadcasters want it to ensure that viewers find favorite and new programming, that they’re engaged and that they return.
The brains behind the bots
We’ve all seen recommendation technologies at work. Netflix, Yelp, Amazon and others use a technique called collaborative filtering that analyzes selections that a person made in the past, and employs that data to make recommendations. Collaborative filtering is also used to recommend new programs based on what other viewers with similar profiles have watched.
Pandora, the music service, and Jinni, an entertainment discovery service, use a different technique, called content-based filtering. Jinni and Pandora developed “genome” projects to map and analyze the characteristics that inform their recommendations. Jinni maps and analyzes categories including mood, style, plot, and setting, and in total uses 2200 different tags to describe entertainment, including things like “uninhibited rivalry,” “fish out of water” and “bittersweet.” Through its Genome project, Pandora uses more than 400 attributes to describe songs.
Discovery: The next generation
Today, discovery technology goes well beyond simple grid listings, suggestions and standard search categories (like “Westerns”). Discovery offerings now let viewers browse and search based on a much more expansive set of criteria, including their mood (I’m feeling nostalgic) genre, location, time-period, plot (nerdy girl gets cool guy), and more. Viewers can create their own custom channel guides built around and across favorite TV channels, streaming video apps and websites. For example, a custom channel line-up might include NBC, UStream, Netflix, YouTube and Facebook.
Some discovery offerings let viewers create customized single or multiple virtual channels, delivering all the titles that, for example, are similar to “24” or “Modern Family”.
The technologies also allows viewers to search across devices and integrate social into the experience, so they can see what friends are watching or recommending on Facebook and Twitter.
Don’t forget the enterprise
Discovery technologies not only benefit viewers, they also provide broadcasters, and publishers and with powerful tools to improve their video business and drive more revenue. Presenting viewers with personalized content is proven to increase engagement time by providing them with what they came for and more —relevant and appropriate entertainment. Case studies routinely show that implementation of recommendation engines, increase viewer engagement upward of 20% and, in some cases, as high as 100%. This, in turn, results in more revenue, in the form of content through advertising, subscriptions or individual transactions.
One Australian news outlet’s recent experience with discovery underscores the value of personalization. The outlet wanted to drive traffic to its web site without investing in new content or building new applications. It incorporated video recommendation technology into its existing video archive and increased views overnight by more than 40%. The result was an immediate and significant increase in ad revenues associated with those views.
In the future, discovery will play a key role in helping broadcasters and publishers earn more revenue from videos while providing viewers with a much more personal experience. Advances may include discovery’s ability to show broadcaster how doubled ad loads impact viewing times, or to suggest the best place to insert a paywall in long-form content.
What lies beyond
As digital programming and viewing devices continue their exponential growth, and discovery technologies improve, demand for personalized TV and video will undoubtedly increase as well. In fact, according to ABI research, by 2018 North American consumer exposure to advanced recommendations engines will reach 75% of pay-TV households on multiscreen services and 55% on set-top boxes (STBs).
Even before 2018, we’ll see many changes in discovery offerings. For example, recommendation engines will focus more on contextual data versus tagged-terms or analyzed categories. These advances will be driven in large part by the rise of mobile devices and on-the-go viewing. Recommendation engines will analyze and interpret data regarding a viewer's’ location, device being used, and time of day. Some will recognize and recommend noteworthy events -- such as breaking news, etc --by location. Contextual data will help differentiate between a viewer in San Francisco on a smartphone at noontime and a New Yorker on a tablet at 6 pm -- even when both have similar profiles or search terms.
Discovery will personalize viewing in ways that will make us feel as though we do have a friendly robot assistant serving up exactly the entertainment we want, when we want it. R2, can you pass the chips and salsa too?