The film industry in the big data age who knows more about the audience who earns more

In 2013, the hit of the card house made Netflix, the online film leasing company, once again a high-profile focus, and the company's investment in the drama based on big data, whether it was a drama set or an actor or director lineup, was underpinned by the user's behavior on the site and the use of data. From the date of development is destined to be favored by the audience.

The application of large data is not only limited to the development of film content, from investment decision, development and production to publicity and promotion, terminal screening, the film industry every link is the possibility of data operation. For film investment production, large data analysis can not only for the film's budget, ideas, actors, and directors provide advice, and can accurately calculate the possible return on investment of the film, from the marketing, promotion point of view, the large data can let marketers know who the audience is, and thus targeted to develop appropriate promotional programs While the movie shows, at what time, in what way to play which movie, also need to be based on the data of the audience. How does the big data play a role in these three links?

The Toupai of the card house based on user data analysis

Big data for investment decisions

Netflix: No one knows the user better than you.

Netflix makes observations every day on 30 million "actions" (including playback, pause, rewind, and Fast-forward), 4 million ratings, 3 million searches, and the time and equipment that users watch during the day for 27 million and 36 million registered users across the United States and the world. In addition, Netflix's films and television shows have been tagged with hundreds of labels, including the film's actors, plots, tone, type and so on. In the past, the labels were used to recommend movies and programs on the site based on the preferences of individual users, and now Netflix is starting to make original content based on user preferences.

Netflix CEO Ride Hastings

Through large data and tag analysis, the company knows that a considerable number of users have already seen Fincher's work, that the films starring Spacey are usually box-office, and the British version of the card house is popular. With these three interest groups, Netflix has found a graphical intersection of Venn, proving that configuring a new version of the card house will be a shoo-in. At the same time, the company also based on a precise algorithm to determine how much investment in such a TV show is reasonable.

In the end, the "card house" that Netflix spends 100 million of dollars is widely praised, and investors have raised their enthusiasm for the company. The company's first-quarter revenue was $1.02 billion, an increase of 18% over a year earlier. In the US market alone, Netflix has increased the number of users by 2 million. Netflix's share price is back to $200 trillion, one of the best performing stocks so far this year.

In addition to the card house, Netflix has invested hundreds of millions of of dollars to produce original series, such as The Hemlock Grove, the stunted development, the women's prison, the children's drama "Snail: Fast", and the "Derek" starring Richie Gevist. Although the hype is not as big as the "card House", the Hemlock Grove attracts more viewers than the former, and "stunted development" may go beyond both.

Amazon Studios directly solicit scripts and ideas from its users

Amazon: Ideas are taken from audiences and used by audiences

Last May, Amazon Studios also announced the start of the original program. Unlike Netflix, Amazon is directly soliciting scripts and ideas from its users. By reading carefully, Amazon eventually screened out 25 shortlisted scripts in 2,700 applications. Last December, the Amazon Studios produced a trial for 6 comic plays. The company will play all 6 pieces in the coming months and use its powerful user base to provide analytical feedback. The audience's comments will then determine which of the films have the most potential to be successful in the series. Therefore, the style of the series and the role of expression in the formal before the launch has been optimized, and has a large number of public opinion effects and audience base.

Studio founder of Liu Yan

Studio: 30 parameters predict the success or failure of the film

At home, for the different aspects of film operation, some data analysis, consulting and other Third-party companies began to appear, for the industry's investors (including Bank Securities, fund companies), film production companies, hospital line companies equivalent types of enterprises to provide film production and investment decision-making basis. Studio, founded by Liu Yan, has a role to play, with 10-person companies mastering 20,000 films, 60,000 artists, 4,000 director data, and a preference analysis of about 80% active users of Weibo.

Studio The idea is that the film's name, lineup, script, schedule, publicity points, theme song, Internet copyright, such as more than 30 parameter variables together determine whether a film can be successful. The company's staff sampled 500 of films released since 1990, found relevant data from the Statistical Yearbook, and then used computers to determine quantitative and formulae. Social networks, such as Weibo, involve actors, characters, and movie endings as variables. Through the analysis of data, studio Digital video works and the behavior of the user association. The "ifilm+" developed on this basis can predict the market performance of a movie with 80% accuracy. After knowing what kind of movie will succeed, studio will be able to make recommendations to producers and directors.

Zhang, Zhang Yimou

Big Data for movie marketing

The film company of the internet age

From the point of view of marketing and distribution, large data can let us know the most direct and effective way to sell the film to whom. Zhang, the founder of the company, was positioned as "the film company of the internet age" at the beginning of its establishment. In his view, the arrival of the "Big Data Age", the use of good Internet data analysis, through a series of accurate positioning to accurately meet the needs of consumers, is the core driving force for the development of the company.

le-Vision Pictures Brothers Company Video Network has done many years of network observation, the user needs have accurate data records. By analyzing the registration information and rating, the company knows the user's career, age, gender, cultural background, preferences. At the same time, the offline staff will continue to the audience for the film research. Online and offline data analysis, coupled with the relevant departments and professional organizations published data, through the integration of these three-party data, Le View gradually set up the market positioning Data system for the audience analysis.

The promotion plan of "small Time" is based on data analysis

Take the propaganda of the small time as an example, through the data analysis, le see the film 40% of the audience is high school students, they are Guo, Yang and other creative loyal fans, is impulsive consumers, 30% is white-collar, to "small time" empathy, is the focus of marketing navigation; 20% are college students, they are non-core consumers, But it can affect other audiences, while the other 10% is the main audience between the ages of 26-35, who need to dispel concerns and expand their outreach. Based on the above analysis, Le Vision plan in the country hundreds of cinemas held "small time" 0 premiere Carnival, held a series of targeted promotional activities. The time and place of promotional activities will vary depending on which type of audience is being targeted.

Zhang Yimou joined to open the new pattern of internet era

And Zhang is clearly not satisfied with just using big data for film marketing. May 28, le video pictures in the Super press release, Zhang Yimou officially joined the music as a signing director and art director. Le vision for this event's positioning is particularly meaningful: "Zhang Yimou joined the music vision, open the new pattern of the internet era." Zhang Yimou said: "I like this position." Under the influence of the Internet, some traditional rules of Chinese films are being rewritten, visible and immeasurable changes are taking place, and a new era has come. "Although the plan has not been announced, the road to film production based on Big data has been opened," he said.

"Stunted development" revitalized on Netflix

Large data for playback terminals

Netflix: Before the user asks

Prior to the original content, the delivery of the technology, user interface, video content and computing equipment integration, seamless streaming is the core competitiveness of Netflix. Netflix's users watch about 4 billion-hour programs over 1000 different devices each quarter. To meet this demand, Netflix uses a large number of video servers scattered around the world, dealing with customer information, video recommendations, digital rights Management, transcoding, and monitoring system performance.

"Stunted development" the latest season (fourth quarter) Stills

When a user clicks on a movie to stream it, Netflix will determine in a flash which server has the movie and is closer to the user, and then the server will choose from dozens of versions of the video file, each corresponding to the various bandwidth, device, and language requirements of different users. At Netflix's headquarters, teams of mathematicians and designers are responsible for studying what people like to watch, and developing algorithms and interfaces to provide them with relevant videos to keep them watching.

Netflix makes an analysis every night to see which programs are the most popular. From two o'clock in the morning local time to five, Netflix fills its servers with all the right programs. The most human-bearing video consumes high-speed flash drives, while others are stored in lower-priced, slower hard drives. Using this predictive model, Netflix robs users of their favorite content before they ask for it.

Not only will Netflix be able to recommend to viewers what programs they might like, but they know how the audience likes to see the content. For example, "stunted development" because the plot complex, deep, humorous fleeting, not suitable for television broadcast, in 2006 had to be off the air. But the TV series is ideal for viewing on the Web, and Netflix has released all of the episodes at once, and viewers can watch the show anytime, continuously or repeatedly, and in this way Netflix has created the second spring of "stunted development". This May, Netflix also launched a new 14-episode peremptorily.

(Responsible editor: Lu Guang)