The Power of Recommendation
The retail video rental business and the pay-TV service provider pay-per-view model have a common connection – they’re both heavily dependent upon new movie releases. Here in the U.S. market, that strategy is now being questioned – as the two largest video rental retailers struggle to remain solvent, and as cable companies unite with media companies in an attempt to revive their VoD services.
In contrast, Netflix key performance indicators continue to outperform expectations. They have recently reported that their 2010 first quarter revenue rose by 25 percent, and gross margins by 10 percent – as total subscribers increased by 35 percent year over year. Netflix total number of subscribers (13,967,000) is just behind two major U.S. MSO’s total video subscribers (even though this is not an apples to apples comparison).
Netflix offers a monthly subscription-based service. Access to the Internet is an integral part of their offering, since customers must go to a Web site to select from a library of 100,000+ DVDs, which are then delivered via first-class mail. Subscribers can also watch movies and episodes of TV shows (currently over 17,000 titles) streamed online as often as they want.
A significant difference of the Netflix service is how it is designed – by intent – to proactively encourage subscribers to choose from the “full catalog” of quality video content instead of just new releases. Netflix focuses on good vs. poor content instead of new vs. old.
Currently, approximately 90% of rentals from video retail stores and pay-TV pay-per-view are new releases, while Netflix new releases rentals are just about 30%. Let’s not forget that rights to new content are often more expensive than catalog content. Netflix has been very successful in promoting its full catalog, and it is thanks to its sophisticated recommendation engine.
Netflix intelligently applies the power of informed recommendations to guide their subscribers to content. Netflix suggestions are based on a number of factors – including member pre-selected content “taste preferences” and individual member’s ongoing ratings, fine grained common themes, movie data, local favorites, and inventory levels. Their approach is focused on finding the content that best suits each subscriber – and by doing so, they’ve effectively changed consumer behavior.
The effectiveness of Netflix’s recommendation engine was well demonstrated during the 2009 Academy Awards ceremony. During the 3-hour broadcast Netflix subscribers added 2 million rentals into their queues. Understandably, “Slum Dog Millionaire” was the most added title in the queue, since it won so many Oscars, but at the same time another 56,000 related movie titles were added through the suggestions of its recommendation engine.
And, Netflix continues to influence change. Netflix subscribers who watched streamed content in the first quarter of 2010 reached 55%, from 48% in the fourth quarter of 2009. Instant video streaming will likely continue to reduce the company’s DVD shipping costs, and further improve its gross margins.
Streaming has become very central to Netflix service as it offers instant gratification for subscribers and cost savings for Netflix. Most of the titles available for streaming are from its catalog instead of new releases.
What’s the lesson learned for traditional pay-TV service providers? That growing broad-based consumer demand and maximizing the utilization of available content libraries is possible by harnessing the power of recommendation engines. Instead of consumers finding the content, content now has to find the consumer. Now, that’s a compelling trajectory to an improved and sustainable return on investment.