When it comes to offering recommendations and personalized services to users sifting through mountains of digital content, online video could learn a few lessons from the music world.
Take, for example, Pandora. The personalized-music-recommendations service has created a business out of automatically tailoring music to user tastes and allowing them to create playlists based on their mood and what type of music they want to listen to at any given point in time. The result: an automated but highly personalized service for music discovery.
Compare that to the world of online video. Here users are largely on their own, left to hunt for interesting videos by relying on editorialized featured shows, related-video algorithms that don’t quite cut it, social media and word of mouth.
Video services with quality recommendation engines have proven that they can keep viewers watching longer, or at least get them interested in content that they might not have known about beforehand. A good example is Netflix: While the service is often knocked for a weak streaming-video selection, it has nonetheless done a wonderful job of keeping viewers entertained by its library of largely long-tail content.
What Netflix and other streaming-video providers don’t do is keep a continuous stream of content flowing, which is in stark contrast to the way most people watch TV. Once a viewer is done watching, he or she is largely left to start the hunt for compelling content all over again.
That’s slowly changing, though, and hints of a Pandora-esque service for video, based on linear personalized videos and recommendations, are starting to emerge. While we’re still waiting for one clear winner, it’s worth examining some of the companies that are working toward a solution to the problem. Let’s take a look at where each of the following companies is in its evolution and what it needs to do to become the killer app for video discovery.
Background: Clicker was originally founded by former members of Ask.com as a search engine for web-video content, but in recent years it’s morphed into a personalized discovery engine for online videos. With Facebook Connect integration, Clicker uses information from users’ social graphs as well as information about videos they’ve viewed, commented on or liked on the site to serve up videos based on personal tastes. With its recent acquisition by CBS Interactive, Clicker has some serious backing.
Pros: Clicker is ultra-easy to use: All it takes is logging in with Facebook to get video recommendations. It also looks at and recommends videos across the entire ecosystem of web-video content, not just web originals or repurposed broadcast TV online. That means viewers will be able to watch relevant videos, wherever they come from.
Cons: While Clicker works very well as a personalized recommendations engine, there’s no way to create continuous streams of videos or video playlists. Also, Clicker’s recommendations are tailored to a user’s overall preferences, but they are not necessarily suited to his or her mood at the moment.
Outlook: With Clicker’s large index of videos to choose from and backing by CBS Interactive, I wouldn’t bet against it.
Background: Founded earlier this year, Cull has developed a way to create smart music-video playlists based on the contextual connections of videos that it has discovered on the web. That means that it draws connections between music videos based on how they’re linked to and grouped across the web. The startup was specifically established to focus on the discovery problem for music videos, seeking to make itself a personalized MTV for the web age. Still in beta, Cull.tv has plans to launch more broadly in the coming months.
Pros: Cull has a great interface for creating and sorting through playlists based on whatever mood or type of music video users might want to watch. Viewers can choose from a number of pre-created playlists or build their own, which are automatically updated based on user interactions with videos. Based on what users have skipped or liked, the site provides better recommendations as time goes on.
Cons: The big con here is that Cull is all about music videos. It could be the killer app for discovery in general if it were extended beyond that subset of videos.
Outlook: Cull could be the MTV of personalized music videos. Unfortunately, it probably won’t extend to more general usage.
Background: Originally founded as an engine for creating personalized, shareable video playlists, Redux is now taking data based on earlier user interaction and applying it to create dynamic playlists for a linear, lean-back experience. That means serving up personalized feeds of continuous videos based on connections that have been built by its curators. Perhaps more importantly, it is not focused just on building this experience for web browsers; instead, it is specifically targeting the experience for TVs and connected devices with a 10-foot user interface.
Pros: Right now, Redux seems to be the closest thing we have to creating a TV-like experience for online video. It provides a premade list of channels that users can flip through, as well as the ability to create their own custom playlists. That gives users a bit of flexibility while also keeping the overall user experience easy.
Cons: Redux is focused almost entirely on web-only content. There’s no blending of web originals with long-form videos from Hulu, for instance. That’s fine if all you want to watch is YouTube content, but it does little for those who want new recommendations for traditional TV or to mix the two together.
Outlook: Redux is working on deals with consumer electronics manufacturers and could soon become one way to browse online video in the same way that viewers watch TV.
Background: Shelby.tv was part of the TechStars New York City startup incubator, during which time the startup developed a web-based app for navigating personalized recommendations for videos based on users’ social graphs. Shelby.tv brings in videos that viewers’ Twitter and Facebook contacts have shared publicly. It then displays them in a full-screen, lean-back environment as a continuous linear feed.
Pros: Shelby.tv adds the element of social sharing to its video recommendations, bringing a word-of-mouth element to the videos that viewers are served up. By doing so, it also helps create a social currency for discussing those videos with friends.
Cons: Shelby.tv is limited by the social networks from which it draws its videos. Perhaps more importantly, there are limited tools for refining recommendations, outside skipping videos a user doesn’t like.
Outlook: Still in its early stages, Shelby.tv faces some serious competition from big players. It will be a tough road.
Background: Launched just a few weeks ago, Vidergy is the newest and least mature of the video-recommendation sites we’ve reviewed. It takes metadata and tag data, as well as YouTube playlist data, to help determine the content of different videos and how they relate to one another. On the consumer-facing side, Vidergy has built an intelligent video-recommendation engine that learns what videos a user likes, based on those he or she skips, and how long the user watches; it uses these factors and others to recommend videos.
Pros: Vidergy enables viewers to choose between different playlists as a way to help surface the type of content they’d like to view during any given viewing session, which is a differentiator over other services.
Cons: Even in early launch, the site design and user experience still clearly need a lot of work. Also, the lack of videos other than YouTube is a huge barrier to wider adoption.
Outlook: Not so good.
Background: YouTube has been working on ways to take user-generated content and make it more attractive for users, which is part of the thinking behind YouTube Leanback. Built to emulate a 10-foot TV viewing experience, Leanback enables viewers to connect and instantly begin watching videos that are recommended to them. It also enables viewers to create custom smart playlists of videos.
Pros: Because it’s Google and YouTube, the service already knows a ton of information about viewers, including playlists, subscriptions and previous viewing data. Right off the bat, Leanback has the potential to offer better recommendations than services that are just starting. YouTube Leanback also has the benefit of reach, money and the development resources that come with being a part of Google, which is a huge advantage over the competition.
Cons: The big downside of a personalized-recommendation and linear-stream product from YouTube is that it only has YouTube content. For users that want to watch other content as well, Leanback won’t be the place to go.
Outlook: YouTube Leanback will likely become a large part of the company’s connected device strategy. That means you can expect to see it on a wide range of devices very soon.