Television revolution in the context of large data

Since 2013, the term "big data" has become increasingly hot. The big data is another disruptive technological change in the IT industry following cloud computing and the Internet of things, known as "new Oil" by Amazon's predecessor, Andreas Weigend. Large data will bring profound changes to the big video industry including television, including industry ecology, content production mode, content evaluation standard and business model.

A split TV data application

Xiao Li is a domestic star satellite TV department head, his job is to lead his team on the channel will play the TV series and variety show to carry out a full range of marketing promotion. The national rankings of the channel are sent by SMS to the relevant leader's mobile phone every day. The depressing is that although his team is working overtime every day, the rankings seem irrelevant, but the lagging boards always hit them.

Xiao Li faces difficulties in domestic TV is more common. In the original SARFT "two restrictions" limited by the TV program pattern, the choice of the ability to play and variety show the creative power directly determines a TV ratings rankings. But in most TV program procurement stage, the selection of artificial is more common. For a long time, TV series purchase relies mainly on the internal professionals and experts to decide whether to broadcast. This kind of judging mechanism is decoupled directly from the market, and the evaluation of TV has formed a circle of "post evaluation marketization, artificial selection before sowing". The procurement phase of the problem, Xiao Li's marketing work has become a futile, at most the icing on the cake.

In the Domestic television field, we successfully imitated the American Nielsen's post evaluation model and system, produced the CCTV Sofres and other data companies, and created a television data industry, but in the program production stage, the evaluation system is far from scientific. The reason lies in the different marketization degree of the two stages. In the post stage, belong to the main advertising market, due to the promotion of a large number of international 4A advertisers, thus establishing a relatively scientific evaluation system, and in the program production stage, the market development level is still far from enough.

Two, the video website Big Data tries

Let's take a look at how video sites in the big video industry buy and produce programs. The US internet giant Amazon has released 14 original TV series screenings (Pilotepisode) produced by its original content team, Amazon UBM, which allows us and UK users to buy free viewing. TV Series Preview set is the production of a project developed into a formal series before the development of the sample set, is the TV series development process in the early stages. Web users watch these screening sets and then vote, and Amazon then decides which programs can continue to be developed, making them a complete TV series, and eventually offered to "Amazon Gold" (Amazon Prime) paid subscribers. In addition, in the initial stage of script writing, Amazon accepts TV scripts online, invites consumers to evaluate them, and then chooses the items to be filmed based on feedback.

If the Amazon studios are only based on feedback from netizens, the video web site, the card house of Netflix, has successfully used big data for program production. "Solitaire", the White House drama, is the first original play of Netflix, the video web site, which has become the most online-video-on-demand episode in the United States and 40 other countries. The exclusive copyright of the domestic Sohu video online The play 20 days later, the number of broadcast over 3.43 million times, known as the United States version of the "Huan biography." Netflix has 29 million subscribers and a powerful database of user viewing habits and taste preferences. Netflix is digging deeper into user ratings, viewing records, and user friend referrals, and even collecting data on pauses or fast-forward to find out what users like about video styles, directors or actors. Netflix is based on these vast user information to determine the content of production. As a result, the "card House" is called the calculated TV series, and its success is the success of large data and large video industry marriage. The production process of the card House completely bypassed the ecological environment of traditional American television, and Forbes magazine rated it as "not just a great show, but a big event in TV history".

Three, large data, reconstruct the TV ecology

Back in the traditional television industry, in the big data age, content providers, television stations, advertisers and data research companies have formed a solid ecological chain began to break. Internet-based video operators such as video sites, IPTV and OTT TV have mastered a large number of user information, which can be mined to expand the industrial chain downstream.

In the upstream content production field, the mode of content production is transformed from traditional to C2B mode, we customize content by understanding user's preferences, points of interest, and user behavior, and provide what users want to see. This also explains the domestic and foreign video websites have entered the content original domain reason. At home, the company set up a music video film, Shanda literature also set up a screenwriter company, relying on "big data" to create TV series.

In the downstream evaluation process, because the video operators have a large and accurate user and viewing data, the original controversial sample sampling mode began to become obsolete. Advertisers, though convinced of the authenticity of Nielsen and Sofres, have begun to adopt precise data from operators. In this way, Nielsen and Sofres sampling data markets will gradually shrink, the data market will be from the sampling mode into the precise mode.

But Nelson is also moving with the times. In 2013, Nielsen decided to broaden the definition of ratings, no longer confined to traditional television networks, and launched a system of ratings surveys for OTT Internet TV and the Microsoft Xbox, Apple ipad and many other screens. Nielsen plans to install new hardware and software statistics tools in more than 23,000 sampled homes, of which only 75% are from traditional television networks. But whether Nelson's advance with the times, or on the basis of sample sampling, can delay the decline of the sample industry is inconclusive, but the demise of Nelson and Sofres, I am afraid, is only a matter of time.

Another area downstream is the advertising market, which is closely related to the viewing data. Traditional TV is the pioneer of free business model, that is, to provide free programs to the audience, and then use the attention of the audience in exchange for advertising and advertising revenue, in the process, the audience ratings as a common exchange of currencies. But the use of big data will revolutionize the business model, with traditional ratings being questioned, and the iron triangle that advertisers, television stations and data makers have formed over the years will be broken.

The most important application of large data is to be able to dig out the intrinsic relation. As early as the 90 's, Wal-Mart by virtue of the global satellite information system, the relevance of the relationship in the basket (harsh basket analysis), can be said to be the originator of large data business. The "Beer and Diapers" story, published in the Harvard Business Review in 1998, has become a classic example of MBA teaching in the world and is widely circulated. The story goes like this: In the 1990s, when Wal-Mart executives analyzed sales figures, they found an incomprehensible phenomenon that "beer" and "diaper" two seemingly unrelated items often appeared in the same shopping basket. The investigation found that the phenomenon appeared in the young father. The ultimate reason is that in families with babies in the United States, mothers tend to take care of babies at home and young fathers go to supermarkets to buy diapers. Fathers often buy their own beers while buying diapers.

Wal-Mart's big data is built on the vast global information system that the retail empire is all over the world, and the large data based on open Internet facilitates the direct application of many industries. Unlike the "beer and diapers" case, it is difficult to find a direct cause for the association relationship in large data, but it does not affect the relevance of the relationship being applied to the business. In the television industry, the information of large data has left the development space for the advertising directional push and O2O mode, thus redefining the TV business model and leaving a great imagination space for the future development mode of TV.

As the largest IPTV operator in the world, the listed company of the radio and television department is also in the layout of large data, exploring directional content, directional advertising, related television and other fields. Baidu recently in the acquisition of PPS, relying on massive search data, through Archie and PPS, the introduction of accurate patch advertising form "A search hundred", such large data in the application of video advertising, will accelerate the ability of its advertising to become present.

        Large data age, data mining is destined to be the killer of video operators, including television stations, who really get the basic data and business development capabilities of large data, who will occupy a high position in the next round of development. Operators with data advantages, such as video sites and OTT operators, will increasingly have a competitive advantage, the market share of traditional TV stations will gradually be eroded, video web site industry oligopoly will be in the traditional television industry to reproduce; for television stations, to establish and improve the strategic position of the data sector, From extensive management to meticulous management, the use of Internet thinking to operate TV, is to deal with competition is the choice.