The New York Times had an interesting article, YouTube’s Quest to Suggest More, that covered their goal to have people stay longer on the site which would increase ad revenue. Compared to other Web sites with similar content (or really almost any Web site) they are way ahead. But compared to that other channel for viewing content, televsion, they are way behind. Users spend an average of 15 minutes a day on the site and they spend about five hours in front of the television.
To increase time spent on YouTube Hunter Walk leads a team of about a dozen engineers, designers and project managers who are fine-tuning YouTube to users what they want, even when users aren’t quite sure what they really want. This is where discovery comes in. One way is to select the 10-15 most appealing videos for a specific user from their library of over 100 million.
The process starts with search. The NYT reported that in November, Americans typed some 3.8 billion search queries on YouTube, more than on any search engine other than Google, according to comScore, a market researcher. But there is a difference. While Google queries tend to be very specific, users often come to YouTube with requests as vague as “funny videos.” This is where discovery can help by providing a range of results that are not simply literal matches.
One challenge is when to anticipate the user might be getting tried of their original topic and proactively offer related content to keep them on site. One way to provide good options suggesting videos that users may want to watch based on prior viewing before, or on what others with similar tastes have enjoyed. The effort requires data-mining techniques similar to those used by and Amazon to make music or book recommendations.
Darwin Ecosystems is also in the discovery business. Instead of offering content in a list format based on your prior behavior or others similar to you, it offers a set of topics arranged in a Scan Cloud™ that correlate with the original search term or, as we say attractor. Then you can hop around to explore the related topics based on your interests. It is an alternative model to trying to read your mind as YouTube and Amazon do. Instead, it gives you a range of choices in an easy interface to allow you to better make up your own mind.