Dialogue with Xie Xiaohu, former global R&D director of Google: Put aside more "should" and "best" when making important choices

In the fall of 2003, Tiger Xie, a computer science PhD student at Princeton, decided to give up his coveted Ivy League degree and various offers he had received, and went alone from New Jersey to Silicon Valley.

He was going to work at a startup company he had heard about from his seniors. Although Tiger had no idea what he would do until he started working, he felt he could take a chance because he had already become a loyal fan of its products without realizing it.

"At the same time, I was offered a job by Microsoft Research, IBM Research and other well-known large companies. But I chose to join Google. At the time, it seemed like giving up living in a five-star hotel and choosing to live in an underground garage."

Tiger recalled that at that time, there were no really good search engines on the market. So when he tried Google on the recommendation of his senior, he suddenly felt enlightened because it could really find what you wanted and it was quite accurate. "Once you use it, it's like taking opium. You can't stop."

From "Fire Chief" to "Change Maker"

It was not until the day he started working that Tiger was informed that he would be working on the advertising backend system. In 2015, Google's advertising revenue exceeded $15 billion, with an average revenue of more than $1,000 per second. "Now it is the core of Google's commercialization, but at that time our team only had four engineers. The challenge that needed to be solved was how to ensure the scalability and stability of the system when the number of users, advertisements and keywords were all in a stage of rapid growth."

"At that time, the system crashed almost every day. We carried pagers with us 24 hours a day. When the system crashed, we would get up to see what was going on. Then we would find and fix the problem in the shortest possible time. Every second of loss meant thousands of dollars in lost revenue, so it was a very high-intensity work environment." Tiger said that this pace lasted for about a year, and after the team's efforts, the system gradually entered a relatively stable state, and we were rarely woken up by the pager at night.

If Tiger's work in Google in his first year can be compared to that of a "firefighter," then starting from the second year, his role became more like that of a "changer."

"In 2004, we first used machine learning in business operations. Previously, the system that determined what kind of ads would be generated when users searched was based on a set of historical data and rules. Since it was a rule, it basically relied on our own subjective judgment. Machine learning uses a completely different approach. You just need to throw a lot of data to it, and then it will summarize the correlation between certain inputs and certain outputs. This reduces labor costs on the one hand, and as the input volume increases, the results will become more and more accurate."

After the launch of this new advertising system, the click-through rate of users increased by 20%, laying the foundation for Google's overall commercialization system today, and is still used on mainstream search platforms. Tiger's team also won the first Founder Award in Google's history for this project. It was his initial work in the Google advertising system team that gave him the feeling of "changing the world" for the first time.

From Engineer to Manager

Different from the promotion path of Chinese programmers, in American IT companies, at the same level, professional engineers usually have higher salaries than managers. "So in the United States, you can actually see a more interesting phenomenon: there is a large shortage of technical managers there, because no one wants to do management, they like to do their own work, and there is also a relatively good room for advancement," Tiger told me. Before returning to China in 2009 to establish the Google Shanghai R&D Center, he had always played the role of an individual contributor. If he had stayed in the US headquarters to work on technology, he might have had a faster promotion channel than now.

However, Tiger, who went abroad in 1998, clearly felt that the changes in China in the past 10 years far exceeded those in the United States. Therefore, when Kai-Fu Lee returned to Google headquarters in the United States in 2008 to lobby for someone to come to China to expand the territory, Tiger quickly decided to seize this opportunity which was more challenging and had more room for imagination.

"Will you feel uncomfortable switching from a technician to a manager all of a sudden?" I asked Tiger.

"Yes, I was forced to do something in a hurry from the beginning. When I returned to Shanghai, I was suddenly given a team and had to start looking for suitable projects and setting up an office all over again. This forced me to learn quickly in a short period of time and then sell while I was learning."

"You have built the Shanghai R&D center from scratch to a scale of several hundred people now. Do you have any tricks in managing the team? Or is it because you are naturally more emotionally intelligent?"

"Haha, my emotional quotient is definitely not very high, but it's better than that of the engineers under me. The advantage of being a manager with a technical background is empathy. Because I am also an engineer, I know how they need to be managed. For example, some people are very persistent. If you directly tell them to do this or that, they will definitely be unhappy. At this time, you might as well give them some space and tell them that I have given you a task. No matter how you do it, just get it done. In this way, they will be more willing to accept it."

From Google executives to local entrepreneurs

In 2010, Google announced its withdrawal from the Chinese market, which was a painful and frustrating thing for all Google teams in China, including Tiger. In order to bring Google back to China, Tiger and his team began to conduct a series of research on the domestic Internet environment.

"As the research deepened, I felt more and more clearly that thanks to the rise of mobile Internet, in the next ten years, the global leaders in several fields including social networking, mobile e-commerce, mobile payment, and smart hardware will emerge from China, not the United States." Tiger revealed that since he returned to China, he has continued to receive invitations from major Internet companies including BAT, but there were only two that really attracted him.

One is Didi, which is now valued at US$16.5 billion, and the other is Xiaohongshu, a Shanghai-based startup that rose rapidly last year.

"Didi's model is very operation-oriented. Technology can play a certain role, but the impact of technology on the entire company is still limited. Xiaohongshu is different. First, its business model is something I have never seen abroad. It is unique to China and can be extended all over the world. Second, it is a product and technology-driven company, and I have more room to play in it." Tiger explained why he finally chose to join the latter.

Tiger said frankly that when he decided to leave Google and join Xiaohongshu as CTO, the most common reaction from his friends was, "What?! What is Xiaohongshu?!"

However, whether he dropped out of school to join the early days of Google, or flew back from the headquarters to the rebuilding Google China, Tiger said that one thing he believed in and insisted on was that at important moments and choices in life, it is better to put aside what others think is "should" and "best", embrace more uncertainty, and follow your own feelings.

So, what new changes will Xiaohongshu have after Tiger joins?

"We are currently doing a lot of things, such as the user portraits and machine learning mentioned earlier. Do you think we can see a significant change in sales today? Actually not. But this helps the entire APP to provide more possibilities and imagination based on a better understanding of users. I think the biggest imagination space for products is what we value. What we ultimately value is not that you buy this product, but that you enjoy using this product, the user experience. Just like Google is doing search, focusing on making it better and simpler, and improving overall efficiency."

Tiger hopes that within three years, Xiaohongshu will become a place where you can find relevant information first when you need to improve your quality of life. "We will provide you with a variety of solutions, not just telling you that we have something on sale, do you want to buy it? Xiaohongshu started out as an overseas shopping guide, which actually provides users with information they may need in the context of overseas travel. In the future, when you are getting married, decorating your house, exercising, or having children, Xiaohongshu also hopes to become the platform that users who demand quality want to visit the most."

At the end of the interview, I added a question that domestic engineers often have about their career development:

At present, many technical talents in China are facing a dilemma: startups have a great demand for TAs when building products in the early stage, but once the product is formed and put into operation, the role of CTO tends to be gradually weakened. I am curious about what Tiger thinks?

"I think what you said is quite pertinent. Indeed, the current situation is like this, which is why I did not choose a company like Didi. I think the voice of technology in such a company is very limited. But for a company like Xiaohongshu, technology will play a very important role in the long run. For example, Google, technology is always the fastest leg. But for many other companies, if the technology is always dragged by the business and products, it will indeed be relatively passive."

“So my advice is that you need to first find a company whose technology can run faster than the team, and then find a position that suits you,” Tiger replied.