As VOD and Internet content expand consumers’ viewing options, search and recommendation capability will become increasingly important. A new research report from Parks Associates reveals several success factors in this area, including the ability to present content from traditional and Internet sources, as well as support for multiple types of metadata and multiple search options.
Searching via show title is the top way consumers find content, the report, titled “The Connected TV and Video Experience: Recommendations, Search, and the User Interface,” found. But a significant percentage of viewers search more generically by genre. When seeking out VOD of TV shows, only a few viewers search via the channel itself, researchers found
When respondents were asked what prompted them to choose content, the most popular answer was “TV ad,” followed closely by “friends/ family.” Other methods–such as movie preview (theater), on-screen guide or Internet, print and public display ads–ranked considerably lower, as did social networking. Answers varied by age group, however, with consumers 55 and older more likely to use an on-screen guide to select content, while consumers age 18-24 were more likely to make decisions based on a discussion on a social networking site.
Consumers’ strong reliance on recommendations of friends and family for content selection suggests they could be open to more sophisticated recommendation engines, which Parks Associates classifies into five categories:
- profile based on viewer-entered preferences
- usage-based engine that makes recommendations similar to content frequently selected
- related content- recommends content similar to what viewer has previously selected, taking into account factors such as subgenres, actors, directors, etc.
- social- recommendations are based on viewing habits of other consumers, which could include narrowly defined groups such as friends and family or broader groups such as a geographic region
- promotional/ provider-driven- content provider defines a set of titles it wants to promote and prioritizes them on search results or displays them separately
The researchers note that recommendation engines have “significant potential” but “must be pursued carefully so as not to overwhelm the user with privacy concerns.”
