Fortune had a good article on the NetFlix prize, arguing that it is a good case study on mass collaboration. Indeed, it is. 40,000 teams competed for the $1 million prize. The competition entailed the development of an improvement in the movie recommendation engine at NetFlix. Why is this so valuable to NetFlix? The business model at NetFlix entails being able to accurately help predict what a customer will enjoy, with a particular focus AWAY from the hit new releases. A focus on new releases requires movie rental companies, such as Blockbuster, to stock huge numbers of a title when it is released, only to then find itself with a huge amount of excess inventory just a few weeks later. Thus, NetFlix would like to rent out a more diverse array of titles. Yet, it needs to recommend movies, including many lesser known and older titles, that customers will enjoy. Customers don't know these movies necessarily, so they must trust NetFlix to help them discover what they will enjoy.
What is NetFlix's leg up on the competition? It's not just better statisticians... after all, this was a public competition. Here is the key: As NetFlix's customer base grows, it's movie recommendation engine improves. Why? The algorithms became more refined as the data about customer preferences become richer and more plentiful. Thus, we have a network effect here. The value to a particular customer rises as the number of NetFlix customer rises, because the recommendation engine gets better! With a huge stable of customers, NetFlix has a huge advantage over rivals because of its proprietary database and algorithms. They can't simply be matched by having better statisticians... you also need the sample size and history that NetFlix has. No one else does at the moment.