Bobadilla, Ortega, Hernando, and Gutiérrez (2013) conducted a survey of 300 recommender system papers and noted the temporal distribution of these as demonstrated in Fig. 2. It highlights the increase of interest occurred after 2006, which was when the Netflix Prize competition was announced. Prior to Netflix, publicly available data for collaborative filtering research was much smaller on orders of magnitude, hence the release of this data spurred a burst of interest in the field which has continued to grow. Fig. 3 shows the the contribution and success of Netflix in the industry, and proved the importance of recommender systems in real-world applications.

Some examples of other industries beyond movies and products recommendation include:

• Medical: Most heavily invested field using these techniques for cancer prediction, user-adaptive education, heart attacks, to chronicle disease management (Hassan & Syed, 2010; Kugaonkar et al., 2012; Trajkovik, 2011; Mazurowski & Tourassi, 2011).
• Education: eLearning applications where using peer evaluations could achieve an acceptable prediction quality as of the tutors to effectively minimize the workload for tutors (Loll & Pinkwart, 2009).
• Social networks: social contextual factors derived from user’s motivation, using individual preference and interpersonal influences (Yang et al., 2011; Jiang et al., 2012).
• Scientific community: finding relevant papers among large archives of articles to recommend users of an online community (Wang, & Blei, 2011).
• Transportation: sightseeing and food recommendations based on GPS data and user’s comments to discover nearby activities (Zheng, Zheng, Xie & Yang, 2010).
• Food: restaurants by recommending varied choices of food to customers (Kumar, Tanwar & Nigam, 2016).
• Human Resources: optimizing human resources for government administration to maximize their effectiveness and delivery of task (Isaias, Casaca & Pifano, 2010).
• Legal: a novel approach for creating quality estimations as an input for intelligent tutoring system for students in the field of legal argumentation (Pinkwart, Aleven, Ashley & Lynch, 2006).

User Experience

The company is looking to refresh its ranking system to improve their personalized user experience as it was revealed by recent news articles from TechCrunch and Variety. In March 2017, Netflix announced it is changing its "5-star rating" system with a "Thumbs-up or Thumbs-down" scheme, like the one used by Pandora, rolling it out globally by the end of April. This drastic shift is based on a large-scale randomized experiment they did in 2016 with hundreds of thousands of members, which resulted in 200% more ratings given by those testing the new binary-like rating system. Judging a show or movie with a "yes or no" simplifies the decision making process of users, allowing Netflix to collect more explicit feedback on what you like and what you don't in place of the previous, ambiguous star rating. If it is easier for users to rate a recommendation, its means more data and feedback for their predictive algorithms (Tepper, 2017).

In addition, Netflix is accompanying this new binary system with a new percent-match feature, which uses their data models to show how likely a user is to enjoy a movie or show before watching. For instance, an item with a great fit for a user's taste will show 98% or 100% match, while the recommender system won't show a match-rating for items with less than 50% match for that user (Roettgers, 2017). Overall, Netflix is in the improvement path; they are indeed striving to benefit from data and experimenting towards more optimal recommendations.

Business Forecast

From a financial perspective JPMorgan chooses Netflix as a top pick for 2017 on 'increased global profitability'. As stated in the article:

"We believe Netflix is on track toward significantly disrupting the linear TV market through strong subscriber growth, content differentiation, and a better consumer proposition," analyst Doug Anmuth wrote in a note to clients. "We believe NFLX sets up as a cleaner story into 2017 with pricing changes behind, revenue accretion from higher ASPs [average selling prices], stronger content, & increased global profitability."

Moreover, the investment Netflix made in their vertical integration writing and producing content has brought returns and more viewers. The company just recorded the biggest quarter in its 19-year history, handily beating Wall Street’s expectations while adding a record 7.05 million subscribers. That’s almost two million more new viewers than even Netflix expected, with a fair number of them overseas. The earnings results capped a banner year that saw Netflix launch its streaming service in over 190 countries one year ago. Already, 47% of its subscribers live somewhere other than the US (Alba, 2017).