Shulou(Shulou.com)11/24 Report--

According to news on January 6, at the end of 2022, Robin Li, founder, chairman and CEO of Baidu, spent an hour reviewing 2022 and looking forward to 2023. Li Yanhong told all the staff: "in the face of challenges, we should seize opportunities and stress innovation."

In his speech, Robin Li stressed his long-term concern, "Technology and innovation." Baidu's R & D intensity and R & D share of total revenue are higher than those of Tencent. We believe in this, although we also feel that we have lost a lot of money now, but we believe that it will have a corresponding return in the future. "

When it comes to technology and innovation, Robin Li has talked about the current hot AIGC many times.

Li Yanhong said: "both AIGC and ChatGPT are new things and new opportunities arising from the development of AI technology to a certain extent, although this opportunity is not so clear." Technology can do this, but there is still a lot of uncertainty about what kind of product it will become and what kind of demand this product can meet. So on the one hand, we are very excited, on the other hand, we also feel a lot of responsibility. "

Talking about the commercialization of AIGC, Li Yanhong stressed, "the real innovation is what kind of social impact this technology did not have in the past, and this impact can only be produced through products and commercial applications. This step is actually the most difficult, but it is also something we have to do."

Li Yanhong said: "over the years, Baidu's investment in artificial intelligence technology, now gradually see a variety of new opportunities. How to turn technology into what the market needs, this is the opportunity for Baidu."

At the same time, Li Yanhong also pointed out many problems in Baidu, requiring everyone to face up to the problems in the business. "to transform technology into what the market needs, we need to be more sensitive and be able to grasp the needs of users and customers more quickly and accurately. We still have a lot of homework to do and a lot of lessons to make up."

The following is a transcript of Robin Li's conversation:

Simple appointment: talking about opportunities, talking about challenges, thinking about new requirements

Time flows endlessly, coming to a new starting point. A few days ago, at the end of 2022, Robin and his classmates gathered in the simple appointment studio for an hour to review 2022, look forward to 2023, and answer students' questions. In the face of the changing environment, Robin also sincerely shares a lot of what he has thought and is thinking about.

The whole scene is full of practical information, although the live broadcast atmosphere is relaxed, the content of the chat is very brain-burning, and the knowledge density is very high, just like a valuable cognitive class, which is worth reading and replaying and reading the report carefully. I believe that after learning, we will have a new understanding of how to play and how to achieve better growth in the new year.

Part1 Review 2022, Prospect 2023

Review 2022

A special year, a rare experience

Shanshan: looking back on 2022, I made a puzzle. The students in Beijing are very familiar with it. One is Xidan in early December, with no car and no one. One is the International Trade Bridge the day before yesterday, heavy traffic, very beautiful.

The year of Robin:2022 is a very unexpected year. In my opinion, these are all life experiences. I lived to be more than 50 years old, and to be honest, I think 2022 is a special year, and it is rare to have such an experience in my life. If something unexpected happens, think of a way to deal with it, and then you can find a place where you can do better, which is also good for people's growth.

Looking forward to 2023

Wanting to do something is the characteristic of Chinese people.

Shan Shan: let's take a look at the outlook for 2023. I would also like to show you some pictures. The first is the meeting of the political Bureau of the CPC Central Committee on December 6, and the second is the Economic work Conference of the CPC Central Committee on December 15-16. In a word, the content of the meeting is, "next year is the year of making every effort to grasp the economy." the third is an article shared by the director group about a man who went to Southeast Asia and saw a lot of opportunities there. In this macro situation, what kind of judgment does Robin have on the macro situation and what new opportunities do you think there may be?

Robin: I actually think highly of our Chinese culture. Chinese people want to do things very much, very diligent and hard-working. The central government often says that our economy is resilient, and one of the key factors is that China's national culture is "I want to work, I want to make money, and I want to live a better life, even if I suffer a little bit." Once there is a chance, a lot of people will really rush up.

In the face of challenges, we should seize opportunities and stress innovation

Shan Shan: there are many opportunities from a macro point of view. Are there any opportunities for Baidu?

Robin: there are both opportunities and challenges. More importantly, we still have to seize opportunities and talk about innovation.

Where does growth come from? In terms of innovation, innovation drives growth. Over the past few hundred years, the growth of not only China, but the whole world has been due to innovation and efficiency. Companies like Baidu still attach great importance to technology investment and innovation, hoping to drive our growth through innovation.

Turning technology into market demand is an opportunity for Baidu.

Robin: over the years, we have invested in artificial intelligence technology, and now we are gradually seeing all kinds of new opportunities. In the past, we talked about the first curve, the second curve and the third curve. In fact, even if we look at a relatively mature business like the first curve, there are still many new opportunities in the past year.

Example: AIGC. After doing AI for so many years, we found that in the past AI tried to understand and imitate people, but now AI can generate content. In fact, we have accumulated technology in this field over the past many years.

Whether the product or the function, how to turn the technology into what the market needs, this is the opportunity that belongs to Baidu. I think we are the best in preparation. Of course, it is doubtful whether we are the most sensitive to the market. This requires us to be more sensitive, faster and more accurate to grasp where the needs of users and customers are, we still have a lot of homework to do and a lot of lessons to make up.

Part2 pluralistic questions, answering questions

Topic1 talks about requirements

Seeing that it is also clearly stated in the third quarter financial report that "effective cost reduction and efficiency improvement has brought about the improvement of the company's operational efficiency", I feel that many of the requirements this year have been put on the improvement of efficiency. What kind of orientation will the company have in terms of operational efficiency next year? What are the requirements for employees?

Robin: we pay more attention to technology investment and pay more attention to the second growth curve and the third growth curve. Our investment in the second and third growth curve, whether in Baidu's revenue or profit share, may be larger. Our R & D intensity and R & D share in the total revenue is higher than that of Tencent. We believe this, although we also feel that we are losing a lot of money now, but we believe that it will have corresponding returns in the future.

Reaffirm the nature of business and study the quality of revenue

Robin: in the process of implementing the concept of "reducing costs and increasing efficiency", I would like to talk about some of the problems I encountered, as well as some situations that I did not expect.

I also said at the General Regulatory Commission in October that many people do not have a deep understanding of the nature of business. Since many of our new businesses are losing money, after a long time, people take it for granted that they should lose money. They will not think about whether this business will make money in three years, five years or no more than a hundred years.

When setting up the OKR for these new businesses, the longest dimension is the year, and what is the revenue growth target in the annual OKR. But how does this income come from and how is the quality? Not to mention ordinary employees, even high-level managers don't think much about this problem. One of our star business, a high-growth business, take a closer look, self-developed products only account for more than 20% of the revenue, and more than 70% of the things are resold. What's the difference between you and a trading company? Trade is also bought here and sold there. How many of us are aware of the fact that so many of us are engaged in R & D and that more than 10,000 engineers spend more than 20% of our income on R & D in contrast to the actual results of our performance? How many people realize that the quality of this income is not high?

Gross margin, operating profit and cash flow are better quality information.

Robin: if you don't look at income, what should you look at? Last time, as I said by the General Regulatory Commission, something of higher quality than income is actually gross profit, not gross profit margin, but gross profit.

The company had the concept of gross profit about 20 years ago. When we did the search business in our early years, part of the revenue came from the alliance. The flow of the alliance was actually bought, and the money we earned had to be shared with the partners of these flows. So whether it is the assessment of the entire company, or the assessment of the corresponding business leaders, when I talk about income, it is called revenue net of TAC (Traffic Acquisition Cost). Because the income calculated by you plus the cost of buying doesn't mean anything. The cost of buying traffic minus the revenue will further explain the nature of our search business.

But later, when doing all kinds of new business, people gradually forget: why do I do this new business, why do I lose 1 billion or 5 billion a year to do this thing. Gross profit is a high-quality information, higher quality than it is operating profit, that is, your operating profit is how much. Although you have gross profit, it may still be a loss to reduce your R & D and various expenses. Short-term "loss is because of rapid growth" is OK, but not in the long run, from the operational level to be able to have profit, this is a good business. These standards, our early search business is satisfied. It is precisely because it is so satisfied that our employees and managers at every level have long forgotten this matter.

Shan Shan: maybe it's because the Internet has been like this for some years, it's all at a loss, but it can still be estimated.

Robin: to see why the loss can be valued behind this-because I believe you will make a profit one day. Even when these businesses are losing money, the gross profit is also very high. The reason for the loss is that they spend a lot of money on advertising and research and development, that is, the nature of the business is actually very high gross profit. For example, Taobao, not net profit, depends on income, the vast majority of its income can be converted into its gross profit. Since most Internet businesses are such businesses, we have long forgotten what revenue means, why we should focus on gross profit and why we should switch to gross profit.

In fact, operating profit is not the last step, the last step is cash flow. Even if there is so much income on the book, in fact, the money is confiscated, it is the money that the customer owes you, and it will not work if the customer closes down or defaults in two days. Many to B companies do not have profit, nor operating profit, but no cash flow, no cash flow, and finally the capital breaks and dies. There are many such companies.

"decreasing layer by layer" is not advisable, improve the position and look at the long term.

Robin: we are not familiar with this logic. People are playing with their lives. If the company wants GMV, I will be GMV. If the company wants DAU, I will be DAU. If the company wants revenue, I will find a way to turn running water into income. Sometimes our understanding of business becomes "diminishing". When I align these ideas with our VP and director, they also agree, but when they fall into a director's OKR, some people may have income in their OKR, and some may not even reflect their income. OKR only says how long it needs to grow. In this way, when he passes it down again, the next level becomes to say, "We don't care about how long the boss wants."

Our company may have CET-5 or even CET-6, and when it reaches the most basic level and the real front-line employees, he simply doesn't understand "Why am I doing this?" or although he does, he doesn't think that "this thing has anything to do with me." As a result, the more people go down, the more they are doing short-term things, and after doing it for a long time, they suddenly turn back and say, "Oh, it's possible to do something wrong, this is not what we want," which is very bad, especially when we are willing to spend a lot of money on it. Senior management is willing to invest for long-term results, you can not make money this year, three years can not make money, five years do not make money also OK, but I hope you make money after seven years. But in fact, when we were implementing it, did anyone go to see what happened seven years later?

I hope that people at every level, especially when you want to go to a higher level, you have to think about this from a higher point of view. We're using OKR now, but once we break it down, we're going to look at those numbers, whether it's the numbers in O or the numbers in KR. Even sometimes no matter what O is, if there is no count in O, you don't look at O and just look at KR, just to complete a number, which is often like this. But more often, in addition to meeting the so-called OKR, we should also jump out of these numbers and take a look at whether the numbers I have taken are useful for long-term business growth and long-term health, and whether they have positive or negative effects.

Align repeatedly against inertia

Shan Shan: this year, the employees of ACG and IDG can actually see that the person in charge of the business is actually asked for a gross profit promotion.

Robin: yes, this has been said many times, but it is really not easy to change people's inertia of doing things or thinking. I don't know how many times I told the same group of people in the conference room. I said that I would look at gross profit instead of income next year, but at the next meeting, they still told me that our income was like this and that of other people's income. how we compare.

Every time I encounter such a situation, I will say it again, and the same group of people may have to tell it five or six times before this inertia can be changed. When you ask him if he understands all this, he actually does it, but he says, "that's what I used to do, so I'm still doing it now." Human nature is like this, "get used to it". So today we spend so much effort to align here, it may not be enough to say it again, but it is better than nothing.

Topic2 talks about Strategy

Q3. After the release of the "New Ten articles", Baidu launched the epidemic index, using the flow of information to combat the fear of the unknown. With the launch of Jingtong Mini Program, Baidu is also the first. These rapid responses to hot spots reflect our social responsibility. I would like to ask Robin, in the post-epidemic era, Baidu how to further reflect the responsibility and responsibility, but also contribute to what aspects of value, what is our opportunity?

It is a great thing for people to have equal and convenient access to information and find what they want.

Shan Shan: after the release of the "New Ten articles", it seems that the number of new cases per day is completely out of line with people's personal feelings, and our epidemic index was still very useful at that time.

Robin: this is exactly the time when Baidu can play a role. I was actually very excited in those days. It is a great thing that our data and our services can enable more people to see the truth in a more timely manner and get information that they cannot get elsewhere. That's what our previous mission said, "give people equal and easy access to information and find what they want." this role is becoming more and more important in an environment full of great uncertainty.

Search is actually countless people using their actions to tell us what the situation is like. If we could put this situation together and present it to everyone, he would not be so panicked-not because of how special his surroundings are, but because the whole environment is like this. Especially when he sees predictions that the trend will calm down in a few days, he will think, "I may not need to store food for a year." This has happened many times in history, when people are confused, do not know what to do, do not know what will happen next, we try our best to get timely and accurate information.

Generally occurs this kind of situation, actually does not have much money to Baidu, namely the flow rises, the income drops. But I think it is not only me, but everyone who participates in this work, every Baidu person, will think that it is a great thing that we can play such a role and be relied on by others at this time.

When others depend on you, there will always be opportunities.

Robin: when others are very dependent on you, there will always be opportunities. The whole search business for so many years, the most essential thing is that when others want to know what is true, when others want to know what the other person said is right or wrong, he came to Baidu, he said, "Baidu said so, so I believe."

HCG's classmates have been making demands these days. At that time, they were faced with "too much demand to do it" because many people were looking for medicine and doctors. But no matter how busy or tired we are, we all keep on working, because there are too many people who need us. It feels good, too.

Q4. Recently, there has been a lot of progress in the field of AI, such as AI painting, AI dialogue (ChatGPT), AI writing. How does Robin view these AI application directions, and what plans or deployments will Baidu have in these directions in the future?

The production of AI technology is very exciting and responsible.

Robin: many people have asked me this question. Not only are my friends in the same industry and field around me, but many people in different fields and across borders are asking me what I think of ChatGPT. I am very glad that such a technical direction that we are pondering every day, it is not easy to arouse such a large public opinion field and attract the attention of so many people.

Both AIGC and ChatGPT are new things and new opportunities arising from the development of AI technology to a certain extent, although this opportunity is not so clear. Technology can do this, but there is still a lot of uncertainty about what kind of product it will become and what kind of demand this product can meet. So on the one hand, we are very excited, on the other hand, we also feel a lot of responsibility. As a company, a business organization, we have such a cool technology, can we turn it into a product that everyone needs? In fact, this step is the most difficult, the greatest and the most influential.

Four types of innovation

Robin: I've been looking at some innovative materials recently, which summarize four types of innovation.

The first is innovation from left to right. From science, from the basic theory, step by step into technology, technology into products, and then products to the market.

The second is right-to-left innovation. There is a demand, then push back to see if you can make a demo or a product; the product can be pushed back to see if such a technology can be realized; and the technology can be pushed back to see what the theory behind the technology is and whether such a thing can be realized.

The third is manufacture with the innovation. That is, the innovations that come naturally in the process of expanding the scale of the products you produce. I think China, as a big manufacturing country, has advantages in this kind of innovation, such as the production of new products, such as TSMC. People used to think that it was a contract manufacturer, but now they find that the threshold is very high and not easy. This kind of innovation is the innovation brought about by manufacturing.

The fourth is institutional innovation. The whole organization is an innovation machine, and its purpose or function is to try every means to produce all kinds of innovation.

In Baidu, we are more or less involved in these kinds of innovation. But how to really produce great innovation, how to enable our mechanism to produce innovation more efficiently, this is a very meaningful topic, and this is what I have been thinking about recently.

Commercialization of AIGC, although difficult to do

Shan Shan: if AIGC, what do you expect from Baidu, do you want us to take the lead in finding out commercialization?

Robin: of course. Innovation does not mean that a technology did not exist in the past, real innovation is what kind of social impact this technology did not have in the past, and this impact can only be produced through products and commercial applications. This step is actually the most difficult, but it is also something we have to do.

Topic3 talks about orientation

Q5. The "end-to-end optimization" mentioned by the General Regulatory Commission takes up a lot of space, but it is not easy to understand. How to understand "end-to-end" here? How to define "end"? Fall on the specific business, how to do is "true end-to-end", how to do is "fake end-to-end"?

Explain "end-to-end", platform or vertical integration

Robin: this concept is really difficult to understand because it involves strategic thinking and development concepts. "end-to-end" is not universal, and some companies say very clearly, "I will be a platform, not end-to-end, I believe in horizontal division of labor."

In fact, not only in the IT industry, but also in the automobile industry. In the automotive field, early Ford did end-to-end, making all kinds of parts, but it used assembly lines to improve efficiency. Finally, it is a horizontal and fine division of labor that subverts Ford. Toyota said, "I don't do a lot of things. If I rely on suppliers to make auto parts, it will be cheaper and more efficient than myself." Finally rely on this to subvert Ford. In the field of IT, Apple is obviously "end-to-end", doing everything, so it makes the phone with the highest end and the best experience. The concept of Android is completely different, his philosophy is: mobile phones can not be made or sold can not be sold OK, I am still a very successful platform.

Robin: so, when we are talking about "end-to-end", we are actually thinking: judging from the entire Baidu business line, whether we should take the road of platform or vertical integration.

If the scale is large enough, the "end-to-end" efficiency is the highest.

Robin: we have to make it clear that there must be enough scale, and "end-to-end" is the most efficient. If your business is very small, you don't have to be "end-to-end", and your efficiency is not necessarily the highest.

Whether it's for MEG, ACG or IDG business, when you think about "end-to-end", you have to think: can I make it big enough? When the scale is large enough, after we connect the chip layer, frame layer, model layer and application layer, we intend to optimize this thing and finally achieve the improvement of efficiency. If I can save 500000 per customer, but the required R & D investment is 50 million. This means that if the number of customers is less than 100, it is not cost-effective, it is better not to do this "end-to-end" optimization.

Whether you want to do "end-to-end" or not depends on whether it is "really effective".

Robin: I used to keep saying that "the essence of innovation is to improve efficiency." if you do this innovation to improve efficiency, this is a good innovation. If what you do reduces efficiency, it's not innovation, or it's not good innovation. Therefore, when we look at their own specific business and "end-to-end", we must make a specific analysis of specific problems, and ultimately fall on "the essence of innovation is efficiency improvement" to see if it really improves efficiency.

For many other companies, it's right to be a platform, even for some businesses of Baidu. Because for them, "end-to-end" is not the most efficient approach.

If you make this judgment, you just want to be a platform, thinking that being a platform is more efficient and can make more money for the company, that's fine. It's just that when I analyze some of our main businesses, such as the ACG cloud business, we are neither NO.1 nor NO.2 in terms of scale. And the idea of NO.1 is to make a platform, set up a platform, and let everyone come here to sing. If we say the same thing: I'll set up a stage, and you all come here to sing. Then why do people come to you to sing? But we can do "end-to-end" optimization-if you are doing biological computing, you can tell your customers that the service you get from us must be the most efficient, the best quality, and the fastest; if you are on autopilot, you can tell your customers to put the cloud here. I have everything from high-precision maps, simulation services and Kunlun chips.

Shanshan: in other words, if you don't have so much ecological dependence on you, don't think you can still be a platform.

Robin: yes, if you do it for a long time and no one pays you, no one wants to develop products and applications for your platform, it's not really called a platform. We still need to think about what problems we can solve. But the reason why I said it was difficult in the first place was that it was not a generalization, and I had to think hard about where the business at hand fits this idea.

Q6, Robin said to find the market, choose appropriate indicators, the General Regulatory Commission also stressed that some non-profit indicators are difficult to reflect the essence of the business. Now that we are making an annual plan, can you guide us on how to see the essence of the business and find the right market and indicators?

Only by recognizing the nature of the business can we find the market.

Robin: the two questions just answered, whether they are "innovation is efficiency improvement" or "end-to-end", have something to do with this question.

I just said that innovation efficiency is from income to gross profit, to net profit, to cash flow, the whole process is to help people understand the nature of the business. "end-to-end" is a further analysis of the nature of what you are doing. When you see the essence, and then go to formulate your OKR, and then put the "number" in it, you will know: why do you put this number, why you value this number, what means you use to do this number is right, and what means you use to do this number is not right.

Each business is different, it is very difficult for us to have a unified rule, so this also needs to give full play to everyone's analysis and judgment ability to do. But I know that in the past, when we made a lot of mistakes, we put a number here, and after a long time, we found that saying, "Oh, it doesn't work if this number is right, and it doesn't really achieve what we wanted to achieve." at this time, we marked the wrong market. For example, when he asked for resources, he always said, "We have to bid for Little Red Books. How much advertising money do you have to invest in Little Red Books a year?" but when he asked him to make something like Little Red Books, he also said, "Oh, there are 10,000 people doing Little Red Books. We only have 100 people doing it, so I can't do that." This is a common way of thinking.

Raise the level of awareness and do the technology that the market really needs.

Shan Shan: questions 5 and 6 and questions 1 and 2 are actually related to the nature of business or business. Maybe front-line students don't have much contact with these things, but it's always good to know something. There are many high-T areas of knowledge that are actually very narrow, and they don't understand them, but it's good to know about them.

Robin: when you understand, you can better understand why we are doing this, and even judge whether we should do it or not. Not only the high T, all the people who do technology hope that the technology he does can have an influence. the technology he does is the technology that the market really needs and that countless people rely on, but very often, whether it is due to the limitation of personal ability or the limitation of the system and mechanism, what we technical students do is actually far away from the market, and most of the time they are self-excited and think they are doing a good job. "I posted the top paper, I applied for a patent, which list did I hit?" but after a year or two, I found that it was not used, or although it was ostensibly used, but when you ask these business students, "how much are you willing to pay for this technology", he will say, "it can be used in vain, let me pay the budget, I don't want to do it."

This kind of problem is actually a big blow to the technician, and it took him a long time to understand that "the technology I make is not what the market really wants." This is very bad, it is a waste of resources for the company, and a great waste of time and energy for these students. It would be nice if we had told him earlier that "what you are doing is not good, what you are doing is wrong, and that is what the market needs."

We often see some people write weekly reports in three colors. Green is good news, red is bad news, and black is normal content. Some people's weekly reports are always good news, how much has been improved here and there, and we will never see what progress we have made that does not live up to expectations. But is it true that the things he is responsible for are always in line with expectations? If you don't say which things don't live up to expectations and where you encounter difficulties, how can your superiors help you? On the contrary, your superiors are blinded by you. He thinks you are doing very well and do not need any help. Sometimes these students are not "intentional", it's just a cognitive problem, and they don't look at it from that height, so I have to say these things to help him stand on that height.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.