Don't underestimate the subversion of the big data on the cultural and creative industries

In the 16th session of the Shanghai Film Festival Forum, the topic of large data has aroused people's discussion. The parties to the media on the big data of the disagreement, the discussion spark splash.

For a long time, the film and television industry is said to have unpredictable returns. The famous director Michael Simino's Paradise movie is one of the most loss-losing films in the history of film, and the Chinese famous director John Woo's wind language has dragged down MGM's eventual bankruptcy.

As the film and television industry belongs to the cultural and creative industries, and the creative people's artistic thinking and as a business person of the producer itself there is a certain conflict, so there is a director is the art of compromise, to maintain business and artistic coordination is also the only way for all successful directors. But is it because of this that from the source to control the script, actor and the choice of film style of science, can improve the accuracy of business returns?

In fact, all the categories of cultural and creative industries have such a similar problem, is a product of the duplication and product quality of the offset problem. Expansion, in the cultural and creative industries now become the growth of the national economy in the background, through large data and cloud computing these new technologies can effectively improve the rate of return on the industry? To attract more capital into the cultural and creative industries?

The recent Netflix investment in the "card House" has been unanimously praised. The card house is not only the most watched episode ever recorded on Netflix's website, but it is also hot in the US and more than 40 countries. The "card House" is an American political drama, adapted from the BBC's 1991 premiere British Broadcasting Company, with a 100 million-dollar investment from Netflix, a video web site. It is because Netflix, after analyzing 30 million of users ' big data on film and television consumption, thinks it is profitable. The scarcity of pure political drama fills the blank of the domestic TV drama consumer market, and the TV drama with a strong director's creation style has more repercussions than the recent two years of several similar political themes, such as Starz's "Big Guy", HBO's "Vice president" and so on.

The card house is the first to completely bypass the traditional TV ecosystem of radio and television networks and cable television, and Netflix will broadcast a quarter of all 13 episodes on the same day for viewers on demand. The wave of the card house has continued.

Amazon is said to be producing at least 11 pilot episodes that can only be viewed on the internet because the company is "in war" with Netflix, competing to exploit people's interest in watching TV on smartphones, tablets and internet television, To expand its share in the field of streaming media playback services. At the same time, Microsoft is making its own TV series, and reports say Google, Apple, Intel and Twitter are also considering developing their own similar products.

From this information can be found, on the one hand, the internet industry for the traditional film and television industry chain infiltration, on the other hand, large data on the film and television production industry. It is no surprise that internet companies have collected a large number of user data, thus becoming the main force of data utilization.

Take the card house as an example, how to subvert the TV production industry chain in the big data age?

The famous director Fincher once took the "card House" the adaptation script, has looked for the United States many television stations, but does not have the one dares to pay, because who also can not say that a 20 years ago the old drama whether still has the market. Netflix has similar concerns, and has conducted a "TV drama Consumption Habits Database" analysis. Eventually, Netflix found that the old play "card house" is still hot on demand, and on-demand the play's user base, also almost with the site Fincher, Spacey fan circle coincide, so decided to invest 100 million of dollars in the replay, and by Fincher director, Spacey starring in the play.

Compared to the traditional ratings statistics only thousands of sample households, "count" out of the "card House" database contains 30 million users of the viewing options, 4 million reviews, 3 million theme search, is a veritable "big data." This data is derived from the data resources Netflix has accumulated over the years. When a user logs in to a Netflix account through a browser, Netflix background technology silently records user location data and device data.

The memory code also includes the user's collection during the viewing process, referrals to social networks, and so on. In Netflix's view, pause, replay, fast-forward, stop, and so on are an act that every day users will produce up to 3000多万个 behavior on Netflix. Netflix subscribers also give 4 million ratings a day, 3 million search requests, and ask about the timing and equipment of the episodes. These are translated into code by Netflix and recorded as content-producing elements.

To find a way to analyze big data, Netflix didn't spend much effort. First of all, tens other users of the site to provide a video of 1 to 5 stars rating, a few years down the amount of data related to more than Bai. To find prospective users recommend a new film and television dramas, the recognition of the audience needs an "algorithm."

But turning huge data into productivity is not easy. For years, Netflix has been recruiting large competitions to broaden its data mining capabilities to improve the accuracy of algorithms. At the end of 2005, Netflix opened a data set and set up millions of dollars in bonuses (Netflix prize) to solicit recommended algorithms and architectures that would allow it to recommend a 10% increase in system performance. This dataset contains more than 480,000 anonymous users who scored approximately 1 billion times for approximately 20,000 movies.

Inspiration for China

At present, the domestic cultural and creative industry has emerged from the initial evidence of large data, such as the Wanda Hospital line to the database as an important core asset, is expected to be established by the end of this year, the membership database will reach 6 million, and Wanda just acquired the United States home line AMC member database number reached 4 million. On the other hand, similar to Tencent, Youku and other video sites have a large number of data based on user accumulation, but also began to try to use Hadoop technology for data mining and analysis. But based on large data can not be too much to enlarge the role of large data, and can not be too narrow to understand the significance of large data.

Large data, the ability to quickly collect data and analyze data in diverse or large amounts of data, remains a fundamental insight into consumer demand. Film and television industry has always been a perceptual industry, so in the understanding of consumer demand performance is not prominent, and many creators in the author's films or art films do not take into account the feelings of the audience, which will cause the loss of training a diverse audience, in the long run for domestic film and television industry development is unfavorable The big data is a scientific and structured thinking to view the film and television industry, which can improve the film and television industry returns.

Now, in terms of discerning consumer demand, several movies since the beginning of this year, "Tae-So", "youth" and "Chinese partner", or a well-known version of the youth novel, or to meet the new era of consumer psychological needs and film new media marketing has a unique, it should be said that the initial manifestation of large data, There are preliminary results in scientific insight into consumer demand. And the current domestic film and television industry in a large number of purchases in recent years more well-known network novel copyright, these purchases are based on the Internet literature consumer groups of data analysis, refined to the consumer group of age, gender, education and other dimensions of factors, greatly improve the scientific.

The so-called big data should not only in the production of film and the purchase of copyright when the role of a link, film companies should be large data penetration of the business of every link. For example, in the production process of film and television production team can through micro-blog, SNS and other channels and potential consumer interaction, the initiative to collect consumer data, while changing the management boundaries of film and television manufacturing team, in the production can be fundamental consumer demand for further optimization of the product.

At the same time in the film product marketing, consumer feedback film and television companies continue to interact with consumers to collect all aspects of consumer information, so as to optimize the consumer's purchase channels and consumption channels, with intelligent push to change the old consumer relationship between enterprises and consumers. For example, the new classic culture in the publishing industry tries to build a private cloud to optimize the product's sales channels, there is a large number of ideas in it. So that the previously less clear product sales process become clear, is a worthwhile attempt to learn.

But it's not too much of a myth. The application of large data, large data is a tool and solution, but can not cure all diseases. Film and television industry and other cultural and creative industries, the author with a certain degree of artistic temperament, but a good product must require the combination of art and business effectively. Consumers often have a blindness to their own needs, in the search for specific products, consumers often do not know what they are looking for what kind of products, such as Henry Ford's famous saying: When you ask consumers what products need, they will say please give me a horse running faster.

Therefore, sometimes based on consumer data analysis may lead to unscientific results, on the other hand, too much emphasis on the analysis of large data will interfere with the creation process, making products neither fish nor fowl. In the creative style is disturbed, the director's personal style is no longer obvious, will affect product diversification, reduce product attractiveness. This process is especially prudent when the consumer is immature and needs nurturing.

The domestic environment is not mature enough to promote film and television investment around big data. As far as Internet video is concerned, the content of Chinese video website is decided by the interest of capital, but the domestic video paying market is not mature, the current main profit mode or advertisement, it is difficult to completely take the user as the center to decide the "configuration" of film and television drama. Only pay TV coverage cost, multi-screen fusion acceleration, video service can break through the category of network computer, become "living room culture", it is possible to realize the custom service by the data.

Second, marketing must be related to content, according to the audience preferences tailored programs, not only can be directed to the associated audience, but also according to the audience positioning "customized" ads. The most critical factor is the domestic data analysis habits did not form, on the one hand, "the evil ratings" "ratings fraud" and other ratings of the wave of statistics, on the other hand, accurate data analysis has not been included in the film and television drama investment costs, often a Shing, not to mention the "calculated" a

The author thinks, the potential of big data is very big, but it needs to be better to embed large data into the product creation process, on the one hand, according to different data dimensions to collect effective data, optimization analysis algorithm, on the other hand, in product marketing, copyright purchase and creation and consumption channels to start, so that large data is not just a more More infiltration of the company's management process into the business assets and management elements.