Portal site to the self-proclaimed "champion"

Four major portals launched the first "Celebration page" of the Olympic Channel flow 2008 Beijing Olympic Games just came to a curtain, four portals staged an Olympic channel flow First battle. Sina, Sohu, NetEase, Tencent, have selected a number of third-party monitoring research institutions of the Olympic monitoring data, announced their own during the Olympic Games during the flow and a number of performance than competitors, Sanna first. The same competition, the result is that everyone is "champion", in the end is the four major portals of the self-play? Or is there a problem with the third party data?

Four major portals:

They say they're "champions" in the Olympics.

Sina throws its own favorable Chinese website ranking network, CTR Market Research, Eric Consulting 3 Third-party Monitoring and research institutions authoritative data: China Internet Association sponsored China website Rankings Network monitoring shows that since the opening of the Olympic Games, Sina Olympic users and traffic two indicators are approaching the 200% of the average value of the four portals, Ranked top of the list for 15 consecutive days, became a well-deserved report champion in the network media, CCTV market Research (CTR) August 9 to August 18 daily to the domestic 128 city netizens "Olympic events Watch survey" results show that through the Sina network to watch the average proportion of users 39.68%, Stabilize the first of the major sites, and Tencent, Sohu, respectively, 35.56% and 22.88%; According to a number of research data, the Olympic Games during the outbreak of major hot events, the majority of netizens preferred Sina to watch the relevant video, and more than 70% netizens think Sina video clear effect of excellence, More than 90% netizens expressed great satisfaction with the depth and professionalism of Sina's report.

Sina has been seized with the "first portal" Sohu also take the Chinese web site rankings, Ctr Market research, Eric Consulting 3 said things, said that the 3 data together show that the first choice for the Olympic users Sohu, Sohu in the Olympic report the entire process has won a comprehensive victory. In addition, Sohu also has cnnic China Internet Network Information Center, Wanrui data, Dcci Internet Data Centre, Analysys International Research Organization, Tsinghua University Media Investigation Laboratory and so on 5 institutions data backing, their data also showed that Sohu during the Olympic Games in the performance of four major portals.

With Sina, Sohu and the mutual War of words, Tencent High-profile throw out the Olympic transcripts, but also in Beijing to hold a high-profile Olympic report "Celebration Conference", posing a full champion posture. Tencent cites data from Nielsen and Alexa's two leading research agencies. Nielsen said that the first time through the Internet to learn the first gold information 18~30岁 netizens, 58% is learned by Tencent, leading the second door of more than 44%; and the global Internet traffic monitoring agency Alexa latest data, said that during the Olympic Games China's four major network portal traffic rankings, Tencent has been the first , during the Olympic Games, Tencent Net Daily daily traffic reached 1 billion levels, set a new record of Chinese internet traffic.

Compared to the other three portals, netease slightly low-key, but also claimed its own Olympic video hits more than 350 million, independent visitors and Page browsing volume of two indicators are in the portal first.

Third-party monitoring bodies:

The website takes the different data to say the matter separately

Four major portals are justified to call themselves first, this "first" how to draw? At the same time, the reporter found that Sina, Sohu Common cited CTR, China web site, Eric Consulting 3 third party monitoring investigation agencies data, the same monitoring agency for different sites of the same incident data results why so conflict? "China News and publication" reporter interviewed Ctr, China Web site, Eric Consulting related people.

Ctr said: "Sina, Sohu uses our two times different survey data, these two surveys of the city scope, methods, etc. are not the same, the two data results are not comparable." Sina released the result is our 128 cities in China to take computer-assisted telephone interview results, and Sohu released the result is our in Beijing, Shanghai, Guangdong, Qingdao, Nanjing, 5 major cities to take the way of the survey results. The 5 most important cities and the other 128 cities of the Internet penetration rate, people's preference for the network are different, the results of the data will certainly reflect the difference. "As to why the two of Sina, Sohu survey selected different urban areas, methods to carry out?" CTR Response: "We have their own research projects, and did not specifically for Sina, Sohu during the Olympic performance during the survey." ”

China's Web site rankings and CTR is similar to the story. Chinese web site Ranking network told the "China News and publication" reporter, we will not take the initiative to provide them with data, we publish the data are on their own web pages, who log in can be consulted. As for the site to get data on how to quote, calculate, this is not what we can grasp. But it is speculated that the results are different because each algorithm is not the same, such as a site may use their own Olympic channel traffic to calculate, and b site may use its own total station traffic to statistics.

During the Olympic Games, Eric was commissioned by Sina to do the Olympic heavy hot event netizens Choose website to obtain information on the investigation. In fact, as a research institution, we all hope to present a relatively objective result of the whole Internet report through scientific research methods. The research itself is commissioned by the client, so the entire data ownership results are theirs, they have the right to carry out some of the results to use, we can not limit the point he must be presented from which angle, "Eric told the Chinese press and publication newspaper reporter," and different websites use the same organization's survey data to show different results, there are two core reasons: first, the investigative agencies themselves in different sites for the selection of different research methods; second, the results of the survey is not for each site is beneficial, but for the market and brand needs, their home in the data interpretation and publication, Tend to choose the part of the indicator data that is good for you. ”

With Sina, Sohu use the same third party monitoring investigation agencies data fight different, Tencent independent selection of international big-name third-party monitoring agency Nielsen data to show authority. It is understood that Nielsen's Web site monitoring techniques for the Web page embedded code, that is, in the Tencent Web page embedded a code, users click once per login Tencent network will be Nielsen's monitoring system recorded. "But Nelson has not been allowed to embed Sohu. Sohu told the Chinese press and publication of the reporter. On the premise of no comparison, Tencent's "Gateway ranked first" results are worthy of exquisite. According to the industry sources, recently, Nielsen on Tencent fraud data to Tencent has sent a letter of attorney. Tencent told the "China News and publication newspaper" reporter, this statement is not true, is a competitor's hype. But when the reporter called Nelson, the relevant people refused to tell the truth, said it would not make any comments, the recent release of relevant statements.

Expert opinion One:

Do not rule out the fraud of monitoring agency data

For the four major portals "the first competition for the Olympic flow" phenomenon, magnanimous consulting President Cheng Tianyu frankly: "Each site needs to say things, will take their own favorable local data." Most of this data is paid for by a website that is bought from a third-party monitoring agency. ”

Wanrui data CTO Yang Hai use the idiom "elephant" to illustrate this phenomenon. Yang Hai that we cannot deny that any blind man touches an elephant, nor does he deny the fact that every blind man speaks, nor does he deny the way the blind find the truth by groping. However, every blind man realizes that the elephants are different, and those who really know the elephants will not agree with any blind person who uses the correct method to get the real facts. In fact, the reason is very simple, the elephant is the same elephant, the method itself is not wrong, wrong in a few blind people do not know the full picture of the time, they are eager to replace the whole picture. This is the objective cause of this phenomenon.

"However, also does not exclude individual monitoring bodies, in the increasingly fierce data monitoring market competition, quick success or by its impact, can not be a sense of comprehensive, objective, development vision and action to truly carry out the mission of monitoring agencies." If the monitoring bodies are such one-sided, subjective, isolated release and interpretation of data, it is no wonder that some sites for commercial interests, intentionally or unintentionally use of data, causing ' the same elephant, different appearance ' this strange phenomenon. "Yang Hai stressed.

Expert opinion Two:

The current sampling method is unscientific

Experts believe that the science of investigative methods is also a problem reflected in the incident. At present, the domestic third party monitoring institutions generally use the sampling survey method for investigation. "Sampling survey can not guarantee the average sample point." "In contrast, web-embedded monitoring is more scientific than a sample survey for the Internet," Cheng Tianyu says. ”

Yang Hai specific analysis: "The disadvantage of sampling method is that it is difficult to ensure that the sample is truly representative of the total quantity, in China, not excluding a considerable number of people to accept the sampling method of acceptance, in particular, many high-end users, more reluctant to spend additional time in the process of receiving sampling." In addition, for privacy, security and fraud-prevention considerations, will make the representativeness of the sample survey is greatly limited. In addition, the sampling survey is to obtain data by subjective selection or subjective description of the respondents, not excluding the subjective description of a considerable number of respondents and the objective reality has a certain gap. ”

"While web-embedded monitoring is an objective monitoring, it does not require respondents to make subjective statements and therefore does not occupy additional time for respondents." Web page Embedded monitoring is a method, it can do local content, specific user groups and specific time period of monitoring, so in a sense, the Web page embedded monitoring methods can do sampling monitoring, but also to do a full amount of monitoring. "Yang Hai said.

"But because China's internet is growing so fast that so many infrastructure projects are too late to complete, the monitoring code used for web-site embedded monitoring is deployed to the Web sites on every channel, for some sites with more complex content structures, The process of the layout of the Web site publishing system automation requirements slightly higher. Many of the more powerful Web sites are able to embed the monitoring code accurately into all monitored web pages in just a few 10 minutes. The operation of embedding code is like publishing Web content, there is no technical threshold, the application of automated publishing technology will greatly improve the efficiency of the whole station code. "Yang Hai thinks.

Expert opinion Three:

The chaos of the entire data industry

In fact, the first battle of the Olympic report flow is only four major portals long hit the "portal first" mouth of a water war innuendo. "The core of this phenomenon is the confusion of the domestic data industry, the industry standards have not yet formed," Cheng Tianyu, "many authoritative data is unreliable." ”

Yang Hai that the data unreliable phenomenon and the overall environment of the domestic Internet and a wider range of integrity of the environment. From the perspective of Internet industry development, network advertising is the main revenue of many websites, and according to the practice of advertising industry, the media in the advertising process is to provide third-party data to the advertising company or advertisers (including media value assessment and advertising effectiveness evaluation).

Yang Hai said that the internet industry itself has a certain new characteristics, but also a high-speed development of the young industry. In the face of such a new industry, the third party data services have a certain lag is inevitable, this lag in addition to reflected in many domestic third-party institutions lack of effective technical means and technical capabilities, but also reflected in the chaos of data standards. Of course, there is also an important reason is that there are some special "PR data" of the "pseudo third party agencies", these institutions in the Internet industry is a serious lack of third-party data, advertisers are not in the market environment to play a "less than" role, but with China's Internet third-party data market gradually mature, Such institutions will eventually be abandoned by the market – as advertisers and advertisers become more mature, and the Nanguo of third party data agencies will get worse in the years ahead.