滴滴CTO张博：我人生中重要的四次选择

回想我前三十几年的人生，几次重大选择都是来了武大之后做的。2001 年，我刚考上武大时学的是自动化专业，我本来以为是造机器人的，结果发现学的是火力发电厂的自动化，我对这个并不感兴趣。

大二的时候，听说国际软件学院面向全校大二学生招生。我当时隐约有一种感觉：计算机能够同时连接几百万人甚至几千万人，我觉得这件事很酷，能很大程度地改变世界。当时我对计算机的了解仅限于聊 QQ、打联众游戏，不确定自己是不是真的喜欢，以及在这个领域能做成什么样。于是我就先“测试”了一下自己，报了个培训班，学了一些计算机网络方面的知识。学完之后我发现自己特别喜欢。

经过一个月的准备，我很幸运地通过了国际软件学院的考试。当时竞争比较激烈，面向全校选拔，最终只有 80 个同学被选上。

转专业第一个学期，我听课非常吃力。许多同学都有一定的计算机基础，我从零开始，要补的课程实在太多。那个学期，我每天白天听完课，晚上就去机房，把书上所有的程序都重新自己写一遍，晚上 11 点熄灯以后，还要打着手电筒躲在被窝里把当天所有老师教的重新复习一遍，再预习第二天的课，提前把不懂的名词全部查一遍。那个学期每天压力都挺大，感觉就像是重新过了一遍高三。我有一个学习方法，就是每个领域就看最经典的那一本书，还会找来英文原版看，把它吃透的话基本就事半功倍。后来学期结束的时候，我从一个计算机小白考到了全院第二名的成绩。

转学计算机，是我人生第一次重大的选择。它带给我两个重要的启发：第一，一定要找到自己感兴趣的方向；第二，认准方向后，要努力到自己无能为力。

非常感谢在武大的四年时光，找到了我自己喜欢的东西并为之努力，可以说这四年决定了我未来的事业跑道。CXO UNION-CXO联盟（cxounion.cn）

一、“我的底层操作系统”

我的人生至今有四次非常重大的选择，第一次是前面说的换专业，从自动化转到计算机专业。第二次是来北京读研究生；第三次是毕业后去了百度做搜索；第四次是 2012 年从百度出来创业，选择了滴滴。

每次要做选择的时候我就问自己：10 年以后会为哪个决定更加后悔？我觉得人生首先还是要做自己喜欢的事。如果你喜欢的事正好也在浪潮之上，当然是最优选择，但如果不是，也不要违背自己的兴趣爱好去做迎合浪潮的事。每一个人本质上还是要追求对自我的满足和认可。

回顾人生几次的重大选择，我的底层操作系统发挥了很大的作用。这个系统是一个圆，上半圆是“不后悔”，下半圆是“要给这个世界留下美好的东西”。

先说上半圆“不后悔”。

其实滴滴是我第二次创业。2012 年我从百度出来第一次创业。因为我发现，当时百度的很多 APP 并没有做更新，但它的用户量上涨的速度非常快。原因是当时大量的手机从 feature phone 换成 smart phone，智能手机开始流行，而且智能手机的价格从 2000 块钱降到了 1000 块钱，当时小米和华为都推出了千元以下的智能手机。基于此，我判断会带来移动互联网的一波繁荣，当时我问自己一个问题：不出来创业会不会后悔？当时的答案还是比较清晰的。

因为团队的原因，第一段创业不到半年就失败了。那段时间没有收入，没有积蓄，还要还房贷，身上的钱只够生活两个月，压力非常大。我一个人背了一书包的方便面，去威海待了一周。这一周我主要问了自己两个问题。第一，还要不要继续创业？第二，创业做什么？我意识到创业依然是我的兴趣所在。那么创业做什么？我去收集了 2010 年到 2012 年所有能收集到创业公司的资料，把这些创业公司按照赛道进行了排列，不同赛道的第一名估值增长速度是不一样的，这是由这个事本身的基因所决定的。

其中有四个关键要素非常重要，那就是大众、高频、刚需和有用户口碑传播的场景。目标用户群是大众的需求还是小众的需求，决定了创业天花板的高低；用户需求是否高频决定了用户的留存；刚需，我对它的定义是创造的价值可衡量、可防御、可持续；当企业的产品或者服务真正触及用户痛点的时候，用户口碑传播能起到的效果会比任何广告都管用。我用了两周时间看了 9 个项目，滴滴是当时唯一一个这四个要素都符合的。

我的底层操作系统下半圆，是“要给这个世界留下美好的东西”。埃隆·马斯克对我影响很大，因为他一直在致力于解决人类最本质的问题，是一个多领域的连续创业者，给世界留下了好多划时代意义的作品。我希望我老的时候，回望一生，也能为这个世界留下一些美好的东西。

二、“创业初心：用科技让世界变得更好“

让我有创业想法的是 2011 年在百度做的一个公益项目，叫“百度寻人”。当时很多大 V 参与了一个“随手拍街面乞讨儿童”的活动，发动网友把街边乞讨儿童的照片发到网上，因为这里面有些孩子很可能是被拐卖的，发在微博上可以让更多的家长看到，解救这些孩子。当时我在百度负责图片搜索，我们已经有了人脸识别技术，我想到可不可以搭建一个平台，让家长把丢失孩子的照片存上来，我们自动抓取互联网海量用户的照片进行比对，如果发现相似性高，就可以第一时间通知家长。CXO UNION-CXO联盟（cxounion.cn）

大家都知道公众的注意力是有衰减曲线的，互联网的热度不会持续太久，为了赶上热点，我们团队熬了一周，终于把这个产品做出来了。我还记得那是一个周六的凌晨 5 点，我用自己的微博发布了这个产品之后就去睡觉了。当时我的微博总共只有四个粉丝，结果一觉醒来，这条微博火了，被转发了 5000 多次，我之前所有的微博加起来也没有这么高的关注度。

就是这个项目，对我的触动很大，以前我就只知道自己喜欢写代码，但“百度寻人”项目让我意识到技术可以让这个世界更美好，第一次真正认识到了科技的力量。

我创业的起心动念就源于此。CXO UNION-CXO联盟（cxounion.cn）

2012 年，我第一次创业失败以后经朋友介绍，认识了滴滴 CEO 程维。他告诉我，滴滴是想解决出租车乘客和司机之间的效率问题。如果大家还记得，当时乘客和司机之间有非常大的信息 gap。乘客经常需要在路边吸着尾气，排着队“抢”的士，而与此同时，出租车司机很可能在没有乘客的情况下开着空车到处转悠找乘客。实话说，听到程维想法的时候还有一些不确定性，但是觉得值得一试。所以我花了 2 天时间打了 30 多辆出租车进行调研，发现对于乘客来说要排队、要等待、要抢车、还经常面临被拒载的情况，对于司机而言每天有大概 30% 的时间是空车，这部分的时间和油费成本都得自己承担。对于大众而言，确实是一个很痛的“痛点”。

所以我想，通过移动互联网解决乘客和司机之间的信息不对称问题，乘客不用下楼就能把打车需求发出去，不用等待、不用在街边吸尾气；司机也可以第一时间获取客源信息，以移动互联网为媒介，将信息传播交换的成本减少到零。这就是现在的滴滴。

三、“在滴滴九年九死一生“

过去九年在滴滴，可以说是九死一生。

创业之初非常艰难，一开始滴滴只有 80 万创业资金起步，我们只给得起 5000 块的薪水，很长时间都没几个工程师愿意加入我们。而和对手的竞争又非常激烈，当时北京另一家竞争对手融资 300 万美元。既要支撑业务的高速发展，又要打技术的地基，确实是压力非常大。因为资源非常有限，一旦我们做错了一个功能，而竞品做对了一个功能，我们就再也追不上了。

2014 年就发生了和快的那场“补贴大战”，当时腾讯和阿里都想在移动支付上发力，而腾讯自己没有在移动支付里面高频交易的场景，就找到滴滴合作，阿里找到了快的合作。我们的补贴规则是，乘客到达目的地之后，用银行卡绑定微信支付，乘客和司机都会获得 2 元补贴。活动之前滴滴每天的订单量大概每天 10 万单，我们乐观估计会涨 50%，所以准备了 40 台服务器。没想到两周下来，用户涨了 50 倍，涨到了 500 万单，我们的服务器扛不住了，一到高峰就宕机。CXO UNION-CXO联盟（cxounion.cn）

此时就是两边技术团队的 PK 了，用户在滴滴和快的之间来回选择，一家服务器崩了，用户就会涌向另一家，然后这一家的服务器也会撑不住，最终拼的就是哪一家的服务器能先稳定下来。那七天七夜的战况是非常惨烈的，我的内心实际非常忐忑。因为能不能打赢，需要多久，心里都没有底，有同事熬到后面隐形眼镜都取不下来，只能去医院取。

竞品的技术团队也非常优秀，到最后真的就是靠精气神撑着，全力以赴，努力到无能为力。最后滴滴用了七天七夜让服务器稳定下来，而对手用了九天。

四、“时代还留给我们很大的空间和挑战”

现在的滴滴当然已经远远超出我们创业时的想象。过去九年，沿着横轴的方向，我们从北京出发，延伸到了全中国 400 多个城市，还进入了拉美、非洲；纵轴方向，开始连接各种各样的交通品类，做豪华车、专车、快车、优享，做电车和电单车，用科技去设计公交调度，提高公交系统的效率，还在不断优化拼车业务。

经过这些年的发展，滴滴已经有了全世界最大的出行网络，在这张网络之下是一份最具价值的交通大数据。CXO UNION-CXO联盟（cxounion.cn）

在我看来，未来 20 年是属于出行的变革时代。希望有更多优秀的人能够和我，和滴滴一起，参与到改变世界的进程中。2021 年我们正式发布了“滴滴未来精英”计划，面向全球高校吸引优秀的科技人才，有专门的实习生计划、博士后工作站等项目。

然而我们依然面临着很多很大的难题，这个时代还给我们留了很大的空间和挑战。滴滴拥有多元化的业务场景，亟待这些高素质、高潜力的顶尖科技人才与我们“一起用科技 D 造未来”，让世界变得更加美好。

翻译

Looking back on my first 30 years of life, several major choices were made after coming to Wu University. In 2001, when I first entered Wu University, I majored in automation. I thought I majored in building robots, but it turned out that I majored in automation of thermal power plants, which I was not interested in.

When I was a sophomore, I heard that the International School of Software was recruiting sophomores. I had a vague sense that computers could connect millions or even tens of millions of people at the same time, and I thought it was cool and would change the world in a big way. At that time, my knowledge of computer was limited to chatting with QQ and playing group games. I was not sure whether I really liked it or what I could do in this field. So I first “test” myself, a training class, learn some computer network knowledge. After learning it, I found myself enjoying it.

After a month’s preparation, I was lucky enough to pass the exam of International Software School

After a month’s preparation, I was lucky enough to pass the exam of International Software School. At that time, the competition was fierce, and only 80 students were selected by the whole school.CXO UNION-CXO联盟（cxounion.cn）

In the first semester when I changed my major, I had a hard time attending classes. Many of my classmates have some basic computer knowledge. I started from scratch and had too many courses to catch up on. That semester, I listened to the class every day, at night to go to the computer room, the book all the procedures are written again. After 11 o ‘clock lights out, but also flashlight hiding in bed the day all the teacher taught to review again. And then preview the second day of class, in advance of the don’t understand the nouns all check again.

Every day that semester was stressful. It was like living senior year all over again. I have a learning method, is to read the most classic book in each field, but also to find the original English to read it. If you understand it thoroughly, you will get twice the result with half the effort. Later, at the end of the semester, I got the second place in the school from a computer exam.

Switching to computer is the first major choice in my life. It taught me two important lessons: first, you must find the direction you are interested in. Second, look for the direction, to their own efforts to do nothing.

Thank you very much for the four years in Wu University, I found what I like and work hard for it. It can be said that these four years decide my future career track.

“My Underlying Operating System”

Up to now, I have made four very important choices in my life. The first one is to change my major from automation to computer. The second time I came to Beijing for postgraduate study. The third time is after graduation to Baidu search. For the fourth time, I started my own business from Baidu in 2012 and chose Didi.

Every time I have to make a choice, I ask myself: Which decision will I regret more 10 years from now? I think life is about doing what you love first. If your favorite thing is also on the wave, it’s best. But if it’s not, don’t go against your own interests and do something that fits the wave. It is in the nature of each person to pursue self-satisfaction and recognition.CXO UNION-CXO联盟（cxounion.cn）

Looking back on the big choices I’ve made in my life, my underlying operating system has played a big role. The system is a circle, the upper half circle is “no regrets”, the lower half circle is “to leave good things to the world”.

First, the upper semicircle “no regrets”.

Actually Didi is my second startup. I started my first business from Baidu in 2012. Because I found that many apps of Baidu had not been updated at that time, but the number of its users increased very fast. The reason was that at that time, a large number of mobile phones changed from feature phones to smart phones, and smart phones became popular. Moreover, the price of smart phones dropped from 2,000 yuan to 1,000 yuan. At that time, Xiaomi and Huawei both released smartphones priced below 1,000 yuan. Based on this, I judged that it would bring a wave of prosperity of mobile Internet. At that time, I asked myself a question: Would I regret not starting my own business? The answer was clear.

The first Duan failed in less than half a year because of the team. At that time, I had no income, no savings, and a mortgage to pay. I only had enough money to live for two months. It was very stressful. I carried a bag full of instant noodles and went to Weihai for a week. This week I mainly asked myself two questions. First, do you want to continue to start a business? Second, what does a startup do? I realized that entrepreneurship was still my passion. So what does a startup do? I went and collected all the data I could find on startups from 2010 to 2012, and I sorted them by track, and the growth rate of the number one valuation varies from track to track, which is determined by the DNA of the business itself.

Four key elements are important: mass, high frequency, rigor, and word-of-mouth scenarios

Four key elements are important: mass, high frequency, rigor, and word-of-mouth scenarios. Whether the target user group is the demand of the public or the demand of the minority determines the height of the entrepreneurial ceiling. Whether user demand is high or not determines user retention; Rigor, I define it as creating value that is measurable, defensible, and sustainable; When an enterprise’s product or service really touches the user’s pain point, user word-of-mouth can play a more effective effect than any advertising. I looked at nine projects in two weeks, and Didi was the only one that met all four elements.

My bottom operating system under the semicircle, is “to leave good things in this world.” Elon Musk has a great influence on me, because he has always been committed to solving the most fundamental problems of human beings. He is a serial entrepreneur in multiple fields and has left many epoch-making works to the world. I hope when I am old, looking back on my life, I can also leave some beautiful things for this world.

“Starting Out: Using technology to make the world a better Place.”

What gave me the idea of starting my own business was a public welfare project I did in Baidu in 2011, called “Baidu Searching for People”. At that time, many V participated in an activity of “taking photos of street begging children”, encouraging netizens to post photos of street begging children on the Internet, because some of these children are likely to be abducted, and Posting them on Weibo could let more parents see them and save them. At that time, I was in charge of image search at Baidu, and we already had face recognition technology. I thought about whether we could build a platform for parents to save the photos of their lost children, and we would automatically capture the photos of massive users on the Internet for comparison. If we found high similarity, we could inform parents at the first time.CXO UNION-CXO联盟（cxounion.cn）

As we all know, the attention of the public has a decay curve, and the heat of the Internet will not last long. In order to catch up with the hot spot, our team worked hard for a week and finally made this product. I remember it was 5am on a Saturday morning and I went to bed after Posting the product on my Weibo account. I only had four followers at that time, and when I woke up, it was very popular. It had been forwarded over 5,000 times. All my previous posts combined did not get that much attention.

It was this project that touched me a lot. Before, I only knew that I liked to write code, but the “Baidu Search for People” project made me realize that technology can make the world better, and I truly realized the power of technology for the first time.

That’s how I got started.

In 2012, I got to know Cheng Wei, CEO of Didi, through a friend’s introduction after the failure of my first business. He told me that Didi was trying to solve the efficiency problem between taxi passengers and drivers. If you remember, there was a huge information gap between the passenger and the driver. Passengers often have to queue up at the side of the road to “grab” taxis, inhaling exhaust fumes, while taxi drivers are likely to drive around empty looking for passengers without passengers.

To be honest, there was some uncertainty when hearing Cheng Wei’s idea, but it was worth a try. Therefore, I spent two days to take more than 30 taxis for research, and found that for passengers, there is a need to queue, wait, grab a taxi, and often face rejection. For drivers, about 30% of the time is empty every day, and this part of the time and fuel costs have to be borne by themselves. For the public, is indeed a very painful “pain point”.

Therefore, I want to solve the problem of information asymmetry between passengers and drivers through the mobile Internet, so that passengers can send out their taxi needs without going downstairs, without waiting or inhaling exhaust gas on the street. The driver can also obtain the passenger information in the first time, and the cost of information transmission and exchange can be reduced to zero by using the mobile Internet as the medium. This is Didi now.

“Near death in Didi for nine years”

The past nine years in Didi have been a close call.

It was very difficult at the beginning. At the beginning, Didi only had 800,000 yuan of start-up capital, and we could only afford 5,000 yuan salary. For a long time, few engineers were willing to join us. The competition was fierce. Another Beijing competitor raised $3 million. It is very stressful to support the rapid development of the business and lay the foundation of the technology. Because resources are so limited, once we get a feature wrong and a competitor gets a feature right, we can never catch up.CXO UNION-CXO联盟（cxounion.cn）

In 2014, the “subsidy war” between Tencent and Kuaidi happened.

In 2014, the “subsidy war” between Tencent and Kuaidi happened. At that time, Tencent and Alibaba both wanted to develop their strength in mobile payment, but Tencent itself did not have high-frequency trading in mobile payment, so it sought cooperation with Didi, while Ali sought cooperation with Kuaidi. Our subsidy rule is that when passengers arrive at the destination and link wechat Pay with their bank cards, both passengers and drivers will get a subsidy of 2 yuan. Before the event, the daily order volume of Didi was about 100,000, and we were optimistic that it would increase by 50%, so we prepared 40 servers. Unexpectedly, two weeks down, the user increased 50 times, rose to 5 million single, our server can not carry, a peak on the breakdown.

Users choose between Didi and Kuaidi. If one server fails, users will flock to the other, and then the server of this one will also fail. The final battle is which server can be stabilized first. The seven days and nights of the war were very tragic, and my heart was actually very uneasy. Because can win, how long it takes, the heart has no bottom, some colleagues get up to the back of the contact lenses can not take off, can only go to the hospital.

The technical team of competitive products is also very excellent, to the end is really to rely on energy to support, go all out, efforts to do nothing. In the end, Didi took seven days and nights to stabilize its servers, while its rival took nine days.CXO UNION-CXO联盟（cxounion.cn）

“The Times still leave us with great space and challenges.”

Didi has certainly grown far beyond what we could have imagined when we started. In the past nine years, along the horizontal axis, we have extended from Beijing to more than 400 cities across China and also entered Latin America and Africa. Along the vertical axis, it starts to connect various types of transportation, including luxury cars, private cars, express cars, premium cars, trams and electric bicycles. It uses technology to design bus scheduling, improves the efficiency of the bus system, and constantly optimizes the carpooling business.

After years of development, Didi has built the world’s largest travel network, and underneath this network is the most valuable traffic big data.

In my opinion, the next 20 years will be a transformative time for travel. I hope more outstanding people can join me and Didi in changing the world. In 2021, we officially released the “Didi Future Elite” plan to attract outstanding scientific and technological talents to universities around the world, with special internship programs, postdoctoral workstations and other projects.

However, we still face many big problems. This era still leaves us great space and challenges. Didi has diversified business scenarios, and we urgently need these top scientific and technological talents with high quality and high potential to “build the future with technology” with us, so as to make the world a better place.

本文由CXO UNION-CXO联盟（cxounion.cn）根据公开资料整理撰写而成，编辑/翻译：CXO UNIONCXO联盟小O。

如需加入CXO UNION（CXO联盟）高管社群，请联系社群小伙伴哦~

免责声明: 本网站(http://www.cxounion.cn/)内容主要来自原创、合作媒体供稿和第三方投稿，凡在本网站出现的信息，均仅供参考。本网站将尽力确保所提供信息的准确性及可靠性，但不保证有关资料的准确性及可靠性，读者在使用前请进一步核实，并对任何自主决定的行为负责。本网站对有关资料所引致的错误、不确或遗漏，概不负任何法律责任。

本网站刊载的所有内容(包括但不仅限文字、图片、LOGO、音频、视频、软件、程序等) 版权归原作者所有。任何单位或个人认为本网站中的内容可能涉嫌侵犯其知识产权或存在不实内容时，请及时通知本站，予以删除。