Text | Fukato Cross
Some things are not on the scale, and they can't be beaten by a thousand catties on the scale.
In the past few years, Internet workers who have encountered non-compete agreements in batches should have a more personal understanding of this sentence. A large percentage of those who signed a non-compete agreement never expected that the piece of paper would one day really cost them dearly. After all, they still have a bit of a B number for themselves, and the company really wants to go online, and the cost can't be calculated.
The legal basis for the non-compete agreement is Article 24 of the Labor Contract Law, which states that the non-compete restriction only applies to "the senior management personnel, senior technical personnel and other personnel of the employer who are obliged to maintain confidentiality". However, the reality is that more and more junior employees are being included in the scope of non-compete.
An article in LatePost earlier this month cited a case in which a cross-border e-commerce company hired about 50,000 outsourced laborers, and a non-compete clause was added to a new contract with the company in February. In the general perception, there is obviously a considerable distance between outsourcing labor and the role of "two highs and one secret" in laws and regulations.
However, in the past few days, a real senior manager and senior technical staff has encountered a similar situation.
The New York Times ran a report on April 19 that a Pennsylvania court had mistakenly released sealed court documents related to the birth of ByteDance. It mentions two contractors who accused Haina International of bringing some cutting-edge search technology to ByteDance without providing reasonable compensation.
Zhang Yiming worked in Kuxun before, and got acquainted with Wang Qiong of Haina Asia, and later Wang Qiong asked Zhang Yiming to start a business to do a real estate vertical search website Jiujiufang. The entrepreneurial experience of Jiujiufang obviously had a direct inspiration for Zhang Yiming to found ByteDance, and the focus of the above-mentioned lawsuit is whether the process involves improper transfer of technology.
The lawsuit that began in the ninety-ninth room
Founded in September 2009, Jiujiufang used to be the largest real estate search website in China. Zhang Yiming's idea of doing Jiujiufang at that time was to use different products to serve different user needs, so Jiujiufang launched a bunch of APPs, including six products such as "real estate information", "handheld rent", and "handheld buying".
Later, when Byte incarnated as an "APP factory", some people traced this practice back to the 99 room period, but in essence, these are actually two completely different ideas.
Like Baidu, Jiujiufang is a vertical search engine for the real estate industry, but it does not produce content itself, but relies on crawlers to scrape public information on the web, such as posts on 58 or Ganji. This method solves the problem of cold start of the website very well, and hits the pain point of the user at that time. Because mobile terminals were just emerging at that time, the industry had not yet figured out what to do with real estate brokerage products.
However, after the initial completion of the start, Jiujiufang did not switch the mode of platform operation, and still did not introduce a mechanism of strong operation and strong audit. Just as it is difficult to distinguish between true and false rental information on major platforms today, the problem of fake listings at that time will definitely be more prominent.
However, it may not be that Zhang Yiming doesn't want to do it, because this is not a problem of the Jiujiufang family, and platforms like Aiwu Jiwu were once in the limelight at that time, but in the end they also disappeared. In Zuo Hui's words, "the consumer Internet is first horizontal and then vertical, and the industrial Internet is first vertical and then horizontal".
The project of 99 rooms is not very successful, and in the case of matrix product layout, there are only 100,000 daily active users by 2011. Zhang Yiming regards entrepreneurship as gambling, since success is a small probability event, it is necessary to increase the frequency and increase the probability of success. In this case, in 2012, he found a new CEO for Jiujiufang, and led several people to found ByteDance.
In this pile of apps in Jiujiufang, what has a profound impact on all ByteDance products later is "real estate information". "Real Estate Information" can basically be regarded as the early product prototype of Toutiao, but the content focuses on the real estate vertical.
Huang He, the first product director of Toutiao, once described the app as follows: "It is to collect and distribute all real estate information, which is particularly popular." Aggregate the information first and then make recommendations. You can even understand it as the real estate channel of today's headlines."
This set of operation mode of real estate information constitutes the skeleton of all ByteDance's products later. Whether it is the first "funny embarrassing pictures" and "connotation jokes", or the later "Today's headlines", they are all based on the series of actions of grabbing, cleaning, aggregating, and recommending content.
At the beginning, Jiujiufang crawled the listings and content information of other websites, which once caused controversy in the outside world about whether it was illegal, and later the copyright issue was also the biggest crisis faced by Toutiao in the early days.
These origins all corroborate the close connection between ByteDance and ByteDance, but it is certainly untenable to conclude that Byte improperly obtained certain intellectual property rights from ByteDance.
The New York Times article did not mention the specific claims of the plaintiffs, and the relevant court documents involved in the case have been resealed and cannot be consulted, but through the combing of Zhang Yiming's entrepreneurial experience and ideas, we can still have a rough judgment on this war of words.
Does Zhang Yiming need "technical assistance"?
According to the New York Times article, the specific allegations made by the plaintiffs are against "cutting-edge search technology." Whether it is the earlier Kuxun or Jiujiufang, search technology is the support of the underlying layer of the product.
Zhang Yiming once summarized the three characteristics of Kuxun search, including the same city search focusing on geographical location, the time-sensitive instant search, and the vertical search that accurately responds to user needs. Zhang Yiming had been in Kuxun for two years at the time, and his position was the chairman of the technical committee.
In other words, at least during the Kuxun period, Zhang Yiming's technical accumulation in search engines was enough to support the operation of a large website.
Jiujiufang was originally split from the real estate channel of Kuxun, and from this point of view, there is obviously no reason to think that Zhang Yiming, who went around MSRA and worked with Wang Xing for a period of time as a technical partner, also needs investors to ask for technical assistance from the United States when he is doing Jiujiufang.
Of course, considering that Haina had already invested in Kuxun before, it cannot be ruled out that "technical assistance" was done during the Kuxun period. But if that's the case, it's quite far-fetched to have anything to do with ByteDance, which was only established in 2012.
"Shendu" tends to believe that in Zhang Yiming's career from Kuxun, Haina may have introduced some technical resources from the outside world from the perspective of post-investment services, and had already negotiated relevant compensation agreements at that time. But when Byte soared into a hundreds of billions of dollars with TikTok and TikTok, the former tech contractor felt it was time to get a piece of the pie, because the pie was too big for a little bit to pay off.
In fact, in the field of writing code, whether in terms of the overall industry atmosphere or actual operability, it is unrealistic for you to engage in a technology monopoly that no one else has. It is never the technology that plays a decisive role, but the entrepreneur's judgment of the trend and the ability to implement the product and iterate quickly.
The Internet does belong to the category of high-tech industries, which can be seen from the generally good educational background of the founders of Internet companies. When the scale of product users expands rapidly, it is often necessary to refactor the original rough software, and once the concurrency increases, the CTO will inevitably be unable to handle it without two brushes.
But at the same time, it is worth pointing out the fact that this is also the most "knowledge equality" industry. In this field, the most cutting-edge technological advances, whether it is the early operating systems like Unix or the recent large language models, the know-how that should be kept secret in traditional industries have been made public. So, in the past two decades, it seems that no Internet company has failed purely because of technical stuckness. The best practices in the industry are there, and the source code is stripped away for you to see, even if you can't recruit people who can't be hired?
On the other hand, from a technical point of view, it is the recommendation engine, not the search engine, that really laid the foundation for the success of Toutiao and Douyin. While both involve complex data processing processes at the implementation level, the goals they aim to achieve are very different.
Traditionally, search engines have required fast and accurate responses because they are triggered by explicit queries by users, and there is not much room for personalization. The recommendation engine mainly predicts the content that users may be interested in from their behavior and preferences, and has a higher degree of freedom and personalization space. On the whole, one is a passive response, one is an active recommendation, one is user-driven, and one is system-driven.
Zhang Yiming recognized the value of recommendation engines early on. At Kuxun, he wrote a small program to let the machine automatically search for train ticket information, and if the condition is triggered, he will notify himself by SMS. When he later recalled this experience, he once said that the paradigm shift from people looking for information to information finding people was his earliest thinking and practice about recommendation engines. Later, Jiujiufang's "real estate information" product gave him the opportunity to practice and strengthen the idea of information distribution based on the recommendation engine.
However, in terms of the technical maturity of the recommendation engine, until Toutiao was launched later, Zhang Yiming was still in a state of "learning by doing".
At the end of 2012, in the office on the sixth floor of Jinqiu Home, Zhang Yiming convened a team to hold a meeting, the core topic of the meeting was "To be an information platform, it is necessary to do a good job of personalized recommendation engine, do you want to start this thing now?"
When he was cheering up the team, he also said: We can't recommend it, but we can learn it. Later, he learned that Hulu's Xiang Liang was writing a book "Recommendation System Practice", and Zhang Yiming also tried to ask him for an electronic version to learn, although Xiang Liang refused on the grounds that the book had not yet been published. These details illustrate two points: one is that opportunities are always reserved for those who are prepared, but there are no people who are completely prepared; and the other is that when Byte started his business, Zhang Yiming and others still had some intention of rushing ducks to the shelves.
epilogue
Zhang Yiming once said that your perception of things is your greatest competitiveness. Toutiao, Douyin and TikTok can get to where they are today, technology is an important factor, but the cognitive change from "people looking for information" to "information looking for people" is the decisive factor. From this point of view, when the content carrier of the Internet switches from graphics to video due to the upgrading of communication infrastructure, it seems destined that bytes will incubate Douyin from today's headlines.
Therefore, as mentioned above in "Shendu", the purpose of the plaintiff initiating this lawsuit is more like a naked and far-fetched act of "grabbing the cake". However, from another point of view, are the increasingly stringent non-compete restrictions of major Internet companies just as naked and far-fetched? Haina International has sufficient resources to deal with lawsuits, while workers facing similar scenarios are in an absolutely disadvantaged position in the game with large manufacturers.
This is worth reflecting on.