The secret behind the boom in Solitaire: Big Data Evolutionary productivity

In the information society, data will become the core resources, like the agricultural era of land, the industrial age of energy.

China is running towards that time, but there is still some distance.

The secret behind the card house

On this site, users produce up to Sanchicodovan behavior every day, such as collection, referral, playback, pause, etc. As data mining technology matures, Netflix is starting to use the data backwards to produce movies.

One of the most popular American dramas of all time is the card house.

The American political suspense drama dubbed "White House Huan" by Chinese netizens is hit by more than 40 countries worldwide. Serious media-whether the New York Times, the Los Angeles Times, or the latest issue of The Economist-are studying its success in important pages.

The producers of the show are neither television nor traditional studios, but a Chinese-like potato and Youku online video-playing website Netflix.

In the early days, Netflix was a well-known online movie rental provider in North America, and its main business was to rent DVDs by mail to make money. However, in the internet age, this profit model is fading. As a result, Netflix turned to online streaming media, but the transition was unsuccessful and has been sung empty by the capital markets.

Like China's video sites, Netflix is looking for a way out, starting in 2012 to launch its own homemade play.

However, in deciding what to shoot, how to shoot, Netflix is a unconventional, offering its own secret weapons-big data.

It turns out that Netflix began with the idea of the importance of data. On this website, users produce up to Sanchicodovan behavior every day, such as collection, recommendation, playback, pause, etc. Netflix subscribers also give 4 million ratings a day, 3 million search requests, and questions about the timing and equipment of the episodes. These are translated into code by Netflix and recorded as content-producing elements. In the early years, the data had been used by Netflix for precision recommendation, and as data mining technology matured, Netflix began using it to reverse the production of the foreground film.

This time, Netflix engineers found that there was an intersection between the likes of the BBC drama, director Fincher David Fincher and the Spacey (Kevin Spacey), a film that would sell if it met those elements.

Netflix decided to gamble, spending 100 million of dollars on a copyright (almost twice times the price of American TV dramas) that was broadcast in 1990 on the BBC TV drama "The Card House", and asked Fincher to serve as director and Spacey as the leading actor.

As it turns out, they are betting that the card house is not only the most watched episode ever seen on Netflix's web site, but it is also hot in the US and more than 40 countries. Taste the sweetness of the netflix,2013 year will continue to launch 4 of homemade dramas.

Listen to the sound of the data

"Snippets" in social media, such as microblogs and micro-letters, are, to some extent, the voice of data in social media.

Netflix is just one of the pioneers of the Nuggets ' data, and China's followers are emerging.

March 2013, Sohu bought the "card house" China Exclusive network broadcast right. And the process itself is an awakening story.

According to the media responsible for the procurement of Sohu Video Rights Film Center senior director Marco, as early as in May 2012, they were in the United States to see a film on the "card House" information, at that time only the theme and lineup of the play, and did not determine the broadcast platform. Sohu did not have the final decision to buy, but left an impression.

By the 2013 Spring Festival, after watching the sample, Marco hesitated: So high-end film, the market will not accept? He had a sensible reason: some internal executives said they liked it, but his wife could not see it at all.

But during the Spring Festival, Marco found that the play in micro-bo, micro-trust friends Circle of some relatively right to speak in the circle of good reputation, and began to ferment. He also noted that some subtitle groups even worked overtime to start piracy.

Some of the network's mainstream elite crowd for the film endorsement, increased Sohu's confidence. Horse can judge, this play's user group characteristic should be: has the experience and the speech power, the high income.

In many video sites, Sohu video has been pushing the self positioning is the introduction of high-quality American opera platform. If the introduction of the play, on the one hand, Sohu "high-quality American opera brand" positioning bonus points, on the other hand, can also be used to attract advertisers to higher prices for stickers advertising.

Balance, Sohu decided to introduce immediately. After the spring festival, they decided to add a temporary purchase, it took only two weeks to complete the process.

March 2, 2013, "card House" in the online Sohu, the next 10 days, the play will not live up to expectations, quickly rushed to the top of the list.

Marco also won the bet. Their decisions, unlike Netflix, rely on years of data mining, just "snippets" of social media such as microblogs and micro-letters, but in part are listening to the sound of data-the sound of data on social media.

Focus on social media

Social media has left a huge amount of data, these traces of the puzzle, developing a secret and even inexplicable links between the different things, such as the Luo concert and the relationship between the Volkswagen car.

Data on social media are the areas that many Chinese data companies are currently digging into.

It has been 6 years since the University of Tsinghua graduated in advertising precision marketing, and he observes that social media's influence on user behavior is growing. Whether it's buying a car, buying a house, a big consumer, or eating and drinking tea, consumers are more likely to seek advice from a friend on social media such as Weibo. In the future, it is possible to become a market as big as a search for the spread of social media. So, in the 2009 years of a Beijing underground garage chat, hit and friends, created a hung Ming social management.

It provides a "social butler" software that helps businesses manage social media such as Weibo and Renren. For example, in the social background of the hung Ming, fans once commented, forwarded, information will be categorized, the administrator can be based on different levels of response and processing, improve efficiency. If there are negative comments, its negative keyword crawling and monitoring of the technology will play a role, reminding the administrator of the priority of recovery processing, to guide. For some enterprises in large, this technology is very practical, because if simply rely on human resources to do this kind of monitoring, high cost does not say, the reaction is also likely not timely, so that negative public opinion fermentation to affect the corporate brand.

However, these are only low-level applications. "Our ultimate job is to help businesses find potential customers, analyze what they have, what they like and how they want to live, and then do targeted marketing programs to promote real trading." "said E.

Hubei to the South weekend reporter cited an example, such as buying a car, some people may be in the micro-blog directly send a message, said to buy a car, grab this data, you can directly to each other to push car ads. But most people may not just say, that's going to look, "the information actually has a lot of dimensions, and what we need to know is what dimensions are relevant to what we want to be clear about, like buying a car."

When they did data mining, they had a surprising discovery: the people who had listened to the Luo live concert were 30% more interested in SVW's cars. Analysis is also justified, such as the car's audience is about 35 years old, home-type, hoping to improve the quality of life of men, And Luo's audience is very consistent. ”

However, this is not necessarily the real reason, the real reason is not so important. In the big Data age, which is popular in the industry, an important insight by Victor Maire Schoenberg, known as the "great Data Age Prophet", is to give up the desire for causation and focus on the relationship. In other words, many times, as long as you know what "is", and do not need to know "why", as long as it is necessary to know the relationship between certain things enough.

Google has a tool called "Google Flu Trends", which tracks the flu situation across the U.S. by tracking the search term data. If a large number of people search for some keywords, such as thermometer, flu symptoms, chest tightness, etc., the system will carry out tracking analysis, the creation of regional influenza map (gray line). The results are broadly in line with the report of the Centers for Disease Control and Prevention in the United States (dashed lines). Source: Google website USA: influenza-like disease (ILI) data is provided by the Centers for Disease Control (CDC). (Zeng Ziying/photo)

"Influence" is the currency

Think about it, also buy an ice cream, you spend 5 dollars and Obama 5 dollars is not the same-"the United States President of the election" is the signboard of the Fortune Weapon!

Companies based on social media such as microblogs are already very much in existence. These social media management tools are targeted at both corporate and individual users. There are even applications that are trying to quantify the impact of individuals to achieve "commercial".

In this regard, the current Klout, which is based on ranking algorithms and semantic analysis, analyzes the activities of users on social networks such as Twitter, Facebook and LinkedIn, and draws a score that can quantify your impact. Factors that affect the level of scores include the number of active fans, the forwarding rate of messages you send, the interaction between original rates and fans, and so on.

Don't underestimate these scores. Klout has a famous slogan: Influence is Money. Think about it, also eat an ice-cream, you spend 5 dollars and Obama 5 dollars is not the same-"the United States President of the election" is the signboard of Fortune sharp Weapon!

Based on this judgment, some businesses are beginning to practise. February 2012, the enterprise software giant Salesforce introduced a service: let the customer company according to customer complaints Klout rating to flexible handling complaints, Klout value of the higher customer complaints will be faster processing, and the company's attitude and service will be more thoughtful. Luxury shopping website Gilt Groupe also began to offer different discounts based on customer Klout ratings.

According to Klout's platform manager, in the future, those with influence will be free to use the airport's VIP lounge, stay in a better hotel room, and get a higher discount from major brand events.

Why are these companies so "snobbish"? The reason is simple: to a certain extent, those who are more influential tend to be more successful in their careers and have more purchasing power. Even if these people have neither money nor power, at least they have a lot of fans. If companies offend these customers, a negative comment may be conveyed to more people and have a greater impact on the company's brand.

At present, Klout is already a star-level start-up company, with less than hundred employees, has received 40 million of dollars in investment, valuations are already more than 200 million U.S. dollars.

Heavy

"The big data in fact the core is not the ' number ', but the ' big ' and ' according to ', the so-called ' big ', refers to the accumulation of data to a certain degree of magnitude, and ' according to ', is to find out the arguments for business decision making reference. ”

But limiting the concept of "big data" to social media is too narrow. In China, the commercialization of big data is still a long way off.

Born in 1981, Hong Times is one of the people who witnessed the development of China's data industry. 2006, he and his peers 闫曌 together to create a admaster (Chinese name is refined Shuo technology) company, the main business is the operation of network advertising. At that time, coincided with the rise of internet advertising, the advertisers and the Internet media urgently need an effective channel, so the model of refined Shuo technology is also very simple, buy a lot of sites in advance advertising location, and then they packaged to sell to advertisers. Hong Times told Southern Weekend reporters that they were in the industry's leading position in the financial and FMCG sectors, but soon they began to change.

About 2009 years or so, with the network advertising in the corporate advertising spending in the larger proportion of business owners began to pay more and more attention to the effect of advertising, a new market-advertising effect monitoring began to appear.

The first transformation of Jing Shuo technology from this point on, they positioned themselves as independent of the advertising and media third-party advertising monitoring body.

Two years ago, as the concept of big data became more and more hot, Jing Shuo technology began its third transition--into the Big data field. "This is what we wanted to do from the start. "Hong Times said that so many years of technology accumulation, they are now finally waiting for the opportunity." So far, Jing Shuo technology has completed the two rounds of a total of 20 million U.S. dollars of financing.

In fact, since the advertising effect monitoring began, Hong Times has been dealing with data.

"We say that the big data, in fact, the most core is not the ' number ', but the ' big ' and ' according to ', the so-called ' big ', refers to the accumulation of data to a certain level, and ' according to ', is to find out the arguments for business decision making reference. "In the view of Hong Times, the Internet has developed for so many years, the magnitude of the data is not a problem, the problem is how to find the" according to "," now the state is, mostly stay in the ' data, no cognition '. ”

According to Hong Times, five years ago, a data analysis company mining a data may take three months, but now, 3 seconds or 0.3 seconds clock is enough. The speed of data mining has increased dramatically, but this does not mean that people's perception of data has improved. ' Many of the clients they work with have a lot of data on them, but many people don't know how to look at the data or where it's worth, ' said Hong. "We have to start with client education first." ”

However, as the value of large data is highlighted, some companies recognize the opportunities. Major internet companies, such as Tencent, Alibaba and so on, have formed their own data mining team. Companies with large numbers of industry data, such as banks, insurance and electricity grids, have also begun to lay out in this area.

"Admaster will do in the future, is the marketing data processing company." "This marketing, not just advertising, is really about ' operation '," says Hong. "Hong Times explained that their advertising effect monitoring, mostly afterwards or in real time, and big future data, is to provide advice and reference before business decisions-just as Netflix uses data to reverse film production," he said.

The flood forecast, the process, will take at least 3-5 years in China.