“If DJI does not succeed in this transformation, it will fall, it will remain stuck there, and gradually go downhill.”
Mystery project
At the annual meeting before the Spring Festival in 2021, Wang Tao, the founder of DJI, told all employees through a remote live broadcast: He used to be too arrogant and thought “this world is incredibly stupid”, but in fact, many things were not understood, but he hurriedly decided.
This low-key and silent company has actually changed a lot in recent years.
In the past five years, DJI has been silently tossing on various new projects related to vehicles. As early as the end of 2016, a research and development engineer in DJI discovered that a colleague around him was recently arrested to work on a “mysterious project”, but after returning, he was “very frustrated”, saying “the project goal is not clear and the team has only a few people” until In 2017, there were only more than ten or twenty people.
At that time, thousands of employees of DJI generally didn’t know what this “mysterious project” was, and they were too lazy to ask. At that time, it was in the golden age of DJI consumer drones (aerial drones). “Everyone was making drones in full swing,” said a former employee of DJI. Until 2017, when he discovered that no one The competitors of the machine were almost dead, only to realize that the industry may have hit the ceiling.
Now, the “mysterious project” DJI vehicle has finally surfaced. DJI claims to be Tier 1 (a Tier 1 supplier) for auto companies, providing “brains” and “eyes” for smart cars, namely L2+ and L3 “highway” and “urban expressway” smart driving and smart parking. Vehicle system solutions and components. Livox, a wholly-owned subsidiary of DJI, can provide lidar. For this reason, DJI invested in one of the few booths equipped with media interview rooms at the Shanghai Auto Show in April this year. On the left side of the booth is the world’s top auto parts supplier “Valeo”, and on the right is the 7-year-old booth. Lidar company “Sagitar Juchuang”.
DJI Automotive at the Shanghai Auto Show in April this year (Source: DJI Innovation Official)
“I didn’t believe it at first when I heard this news, but after various verifications, the responses I got show that this matter is true.” A “hardcore” who has bought all consumer drones from DJI in the past 9 years Fans” said 36 krypton. This is undoubtedly the result of DJI’s secrecy of the project in the past few years.
“The style of DJI is that you don’t say it, and when you say it, it’s when you can buy it.” It’s just that internal discussions are quite slanderous-starting in 2018, every time the year-end awards are issued, some employees in the camera department We will discuss in private: “Our department has been making money, but why the year-end bonus has not risen? Is it because the company’s overall efficiency is poor, or the boss gives us all the money we make to the automotive department that has not made money for a long time?”
From Wang Tao’s perspective, if you want to return to high growth and invest heavily in the automotive business, it is a decision that must be made — —
DJI, established in 2006 from the student dormitory of the Hong Kong University of Science and Technology, is already an “old company.” After DJI released the world’s first aerial camera “Elf” Phantom in 2012, its revenue and profits have grown rapidly year after year due to the explosion of drones: According to DJI’s public data, its revenue growth rate was still as high as 2017. 60%, breaking through 17.57 billion yuan; net profit soared by 122.8% to 4.3 billion yuan.
But this “growth myth” came to an abrupt end in 2018, and DJI no longer announced its performance.
36 Krypton learned from investors that DJI’s revenue and net profit only grew in single digits in 2018 and 2019, and it will not achieve a year-on-year revenue growth rate of 30% (up to about 24 billion yuan) and 33% until 2020. % Of the net profit growth rate (up to about 6.4 billion yuan).
The observation of many drone users in the past two years is that since the flagship drone Mavic 2 was released in 2018, there has been no announcement of Mavic 3 for 3 years. “It used to be like dumplings. There were several press conferences a year, all of which were heavyweight.” A drone player recalled 36 krypton.
Several DJI’s former employees questioned an incident. At the beginning of 2019, DJI issued an announcement stating: “Due to supply chain corruption in 2018, the average purchase price exceeded the reasonable level by more than 20%, and it was conservatively estimated that it caused more than 1 billion yuan. The loss.”-These resigners suspect that their family scandals and self-exposure to corruption are trying to find reasons for the “not growth” in 2018 and give investors and employees an explanation?
Realizing the bottleneck of growth, Wang Tao has been looking for new directions. From 2014 to 2016, in addition to consumer drones (which accounted for 85% of the company’s revenue in 2017), he continuously expanded three product lines: agricultural plant protection and industrial application drones, handheld imaging, and education. Sources told 36 krypton that DJI had also incubated a social APP “as usual” and has now ceased operations.
In 7 years, compared with DJI’s consumer-grade drone business with a revenue scale of 10 billion, none of the three new businesses can afford to “rebuild a DJI”. 36氪 learned from many industry professionals:
Handheld imaging products such as sports cameras and handheld gimbals have annual revenues of billions, but they are still in fierce competition with rivals such as Insta360, Zhiyun, and GoPro.
The revenue of agricultural plant protection machines last year was about 2 billion (the domestic market accounted for more than 50%), and the industry drone revenue was 1.2 billion (different classification methods may lead to data differences), but agriculture is a chore, and industry drones Although the annual growth rate is rapid, in the security and police use, electric power inspection and other subdivisions, it is necessary to meet the various differentiated customization needs of customers, and it is difficult for consumer drones to use a standard product “winner takes all” situation.
Educational robots can sell for hundreds of millions every year, and holding RoboMaster requires tens of millions of investments every year. The contribution to the company’s benefits will take time to test.
For DJI, which is valued at hundreds of billions, only the trillion-level market of “automobiles” is big enough to be the second engine of this “manufacturer of tens of thousands”.
Although well-known and wealthy, started quite early, and is familiar with hardware manufacturing and machine vision technology, DJI’s road to “building a car” in the past five years has been a tortuous one. It has tried almost all mainstream paths: it has invested in foreign supercar companies, established a joint-venture autopilot company, has been a first-tier supplier of smart driving for car companies, and also sells lidar… Although it has not yet manufactured a complete car, it does not Few people think it should be done.
“If DJI does not succeed in the transformation this time, it will fall, it will stay there for the rest of its life, and it will gradually decline. This is the case for enterprises, and they are willing to bet. IBM will fall overnight, and so will Nokia.” An automobile industry person familiar with DJI expressed concern about 36 krypton.
In the past two or three years, DJI employees have undergone many organizational structure adjustments. “Everyone’s mentality collapsed a bit, and they had to deal with projects and personnel.” A former DJI employee told 36Kr in 2019. Before and after, I received an open letter written by Wang Tao every few months in my mailbox, which contained a lot of “self-reflection”. For example, reflection on the past caused a lot of unnecessary costs because of distrusting colleagues. “He rarely did this before 2018, because the company went smoothly.” The former employee told 36Kr.
DJI builds a car, it’s a story about how to leave the comfort zone.
Outside the comfort zone: Intelligent driving “difficult to make friends”
36氪 learned that in 2016, Wang Tao wanted to be a Level 4 autonomous taxi (Robotaxi). The track was very hot back then, and it has not suffered actual setbacks and has been questioned. At that time, Waymo had just separated from Google to do Robotaxi, and a group of Baidu Meiyan employees had just left and were preparing to start Robotaxi companies such as Xiaoma Zhixing, Roadstar.ai, and Wenyuan Zhixing.
Although Wang Tao is a hardware tycoon, doing Robotaxi is “climbing the Everest of artificial intelligence” and requires strong AI and algorithm talents, but Wang Tao did not find the right person to help him lead the team at the time.
DJI has never worried about hiring people on the drone site before, because hundreds of drone manufacturers around the world are its subordinates, and DJI’s offer is the best. Several former employees of DJI told 36Kr that Wang Tao felt that “excellent people will come to me, and those who do not come are not good”, “My products are awesome, and it must be your problem if you don’t buy them. “.
However, when entering the autonomous driving industry, there are many giants and talents are scarce. The “big cows” in the industry either set up their own companies or are snatched by the giants. What’s more, 36 krypton learned from DJI’s headhunting company and several former employees that DJI rarely gives newcomers stock options and has not announced a listing plan. This may cause DJI’s attractiveness to senior talents to plummet.
The Robotaxi team failed, and Wang Tao had no choice but to retreat. At the end of 2016, he began to consider the development of a Level 2-Level 4 intelligent driving solution for car company customers, allowing Hong Kong University of Science and Technology’s student Shen Shaojiao to lead the team-Shen Shaojiao is DJI The chief roboticist and an assistant professor in the Department of Electronics and Computer Engineering of the Hong Kong University of Science and Technology-established the “Mystery Project” DJI Vehicle.
Shen Shaojiao is not perfect, but may already be the best choice for DJI. Sources familiar with DJI told 36Kr that Shen Shaojie lacked engineering experience, and did not “all in” DJI, and needed to allocate time and energy to do academic work at HKUST. But on the other hand, the artificial intelligence and drone circles mostly praised Shen Shaojie as an “academic model.” The Department of Electronic Engineering and Computer Science, Hong Kong University of Science and Technology, where he is located, also sends a large number of masters and doctors to DJI every year, and Shen Shaojie and Wang Tao met. For 10 years, he is one of the few people who can get along well with Wang Tao and trust each other.
36 Krypton learned from sources that DJI’s previous competitive strategy of “low-price volume” used in drones and handheld imaging products has also been applied to the intelligent driving business:
To “volume”, the millimeter-wave radar selected by DJI’s in-vehicle intelligent driving solution will basically be used by car company customers at a cost price or even lower than the cost price around 2019. To reduce costs, DJI Automotive has removed the original millimeter-wave radar production line in the last six months and transferred it to the factory for production. The purchase of millimeter-wave radar chips has also begun to switch from foreign suppliers such as Texas Instruments and Infineon to domestic brands. “The performance will be lower, but the local after-sales service will be better.” At present, Baoneng Automobile Qoros 7 is using DJI’s forward multifunctional millimeter-wave radar.
Also to reduce costs and power consumption, DJI has also reduced the computing power of the intelligent driving domain controller to 20 Tops (upgradable to more than 100 Tops), which means that models with a price of less than 150,000 can also be used. This has formed a separate line with Huawei’s strategy: the central supercomputing platform provided by Huawei to BAIC Jihu Alpha S has a computing power of up to 400Tops and 800Tops, and the price of the vehicle is more than 388,900.
“Huawei does not want to make cheap ones, to fill the profit gap of mobile phones. But DJI is more like the kind of company that advocates technological equal rights’. It has to make them cheap so that more people can use them.” Autonomous driving executives mentioned 36 krypton.
“Low cost” has fatal appeal to car companies. Moreover, during the 36Kr visit, DJI Automotive’s “soft and hardware full-stack R&D capabilities” and DJI’s “mass production capabilities” proven on drones were recognized by many autonomous driving veterans. They believe that if DJI integrates these three capabilities, it will be enough to “kill” most of its peers.
But the strange thing is that, apart from the officially announced new Baojun brand of SAIC-GM-Wuling, DJI has no designated customers for other leading car companies, and the so-called “cooperation” with Volkswagen circulated in the industry has not disclosed details. Huawei, which is also a “cross-border” first-tier supplier, is currently selling mass-produced models in cooperation with Xiaokang Cyrus in its flagship stores. It also announced that it will launch smart electric vehicles in cooperation with BAIC, Chongqing Changan, and Guangzhou Automobile at the end of the year.
At the Shanghai Auto Show in 2021, the new Baojun E300 modified KiWi EV with “Powered by DJI” printed on it (Source: DJI Innovation Official)
“In the beginning, most of the members of the vehicle were transferred from the drone team. Later, they did not dig into the big cow’ and directly took a set of mature solutions. Our technology may not be considered the best. There are still some better technologies in the industry. Choice,” a former employee of DJI told 36Kr that they had tested products with several leading car companies around 2019, but failed to cooperate successfully at that time.
Not only is it difficult to recruit people and technology, but it is also outside of DJI’s comfort zone to become a “B-end” car company customer and become a “Party B”.
“Newcomers entering the automotive industry and getting customers are a very slow process. Strong ability does not mean they can do business. Because the demands of both parties are different. The stronger you are, the harder you are, and the On the contrary, the more cautious. DJI was too easy to make money before, and the requirements for profitability may be higher, and small customers may look down on it.” An autonomous driving industry executive mentioned to 36Kr.
A former employee of DJI said that when DJI used to make drones, it was far from competing products. It could sign some tough terms with suppliers and partners, but now it is the first-tier supplier of auto companies. Business is not the earliest, nor is it the best, but it will inevitably continue the previous arrogance.
When faced with large B-end customers, such as Tencent, to sell corporate services, Ma Huateng and Tang Daosheng have to come out on the platform in person. But Wang Tao is more like a “technical nerd”, low-key and reserved, and likes to hide behind the scenes: In the 15 years since the company was founded, Wang Tao has rarely accepted media interviews, especially in the past three years. This makes his “popular image” remain in the few public photos that were rare three to five years ago: wearing a gray peaked cap, black round glasses, and a small beard on his chin.
Not only does he not like to see the media, but Wang Tao is also not happy to come forward to receive visits from government officials. In the 2018 round of 1 billion US dollars bidding and financing, only the shortlisted shareholders met him at the final meeting, and he is usually the assistant to president Wei Yiman. Others are responsible for financing matters. Even DJI Automotive’s debut at the 2021 Shanghai Auto Show and the large-scale national drone industry forum is also represented by President Luo Zhenhua and corporate strategy director Zhang Xiaonan.
The head of the company keeps a distance from politics, capital, and the media, which can certainly save a lot of trouble, but it will also leave the impression of “unfriendly” and “solitary” on the partners.
“If I really cooperate, I am really worried about the after-sales service. If I call Wang Tao, he will definitely not answer. If you don’t answer the two, I will definitely put him on the blacklist.” An OEM Gao Guan expressed doubts about 36 krypton.
“I can’t imagine DJI cooperating with SAIC and FAW. State-owned enterprises have a lofty attitude, and DJI is naturally proud.” A person in the auto industry mentioned that DJI is not a company that can serve people. Its product logic is to help customers define what you need, just like Apple.
Moreover, among the solution providers, there are not only startups with earlier start-ups and stronger algorithms, but also experienced suppliers such as Bosch and China, as well as giant rivals such as Baidu and Huawei.
Wang Jun, President of Huawei Smart Car Solutions BU, once announced in a high profile that this year one billion U.S. dollars will be invested in research and development, with more than 5,000 R&D personnel, including more than 2,000 autonomous driving teams. “This ruthless remark makes any start-up company think that Huawei is going to dig me!” an executive in the autonomous driving industry told 36Kr.
On the contrary, DJI Automotive’s description of itself is very brief: “As of the end of 2020, there are more than 500 core R&D personnel and more than 200 engineering teams”, and it has not announced the future R&D investment and recruitment scale.
An episode is that in addition to providing smart driving solutions to car companies, DJI also established an autonomous driving company “Fengjiang Intelligence” at the end of 2017 with Dongfeng Asset Management, the investment platform of Dongfeng Motor, to develop Level 4 limited farmland scenarios. Class autonomous driving agricultural machinery. But at the end of 2019, DJI withdrew all 20% of its shares.
Sources told 36Kr that DJI’s divestment was mainly due to “different styles of the two partners” and “DJI still likes to do things that it can control 100%.” But this also means that DJI’s autonomous driving puzzle has been lost again. piece.
“DJI is not willing to make friends, nor does it have friends,” another source who knows DJI mentioned to 36Kr. “In the past, DJI has worked well with DJI, almost none. This proves that Wang Tao can’t cooperate with others. Almost no one can make money with him, including his internal employees. DJI is not in a state of sharing benefits with others.”
Several former employees of DJI told 36Kr that this style has also led to a climax of resignations between 2018 and 2020, including a large number of executives who helped Wang Tao “fight the country”, such as Wang Mingyu, vice president of research and development, and sales Vice President Xie Jia, Vice President of Sales Yuan Shujian, Chief Scientist Wu Di, Founder Wu Xumin of Agricultural Plant Protection Machine, Hong Xiaoping, Head of Lavo Technology, a wholly-owned subsidiary of Lidar…and many other vice presidents, North American industry application executives, one Batch star product managers, etc.
An investor in the enterprise service field analyzed 36 krypton and said that although DJI lacks B-side experience, it is good at “technology-based research and development of system architecture”, and it actually has an advantage in serving B-side. Whether it can be a good first-tier supplier in the future depends on whether DJI has a strong will and motivation to let go of some of the past “characters”, introduce and trust new people, and build the system-for example, learn to listen “from the bottom up” And absorb various “non-standard” needs of B-end customers, learn to share benefits with employees, and learn to balance the interests of direct sales and agents.
DJI’s Expert Methodology
In addition to providing a complete set of intelligent driving solutions, DJI also began to develop the core sensor “Lidar” in 2016. To this end, Wang Tao recruited Hong Xiaoping, an alumnus of Hong Kong University of Science and Technology, who was responsible for sensor research and development in Honeywell (Silicon Valley) in the United States, to lead the team, and set up a wholly-owned subsidiary, Livox. However, in the eyes of many industry insiders, DJI’s lidar solution is a “risky move.”
A few months ago, when Xu Han (pseudonym), a researcher at a car company, got the Livox car-level lidar prototype, he was in a dilemma: “It’s so cheap (less than 10,000 yuan), but the point cloud is so weird. It takes time and cost. It may not be successful to do algorithm adaptation.”
Xu Han was not the only one who hesitated about Livox. An autonomous driving veteran also complained about 36 krypton: “The lidars of other manufacturers are linear and repetitive scanning, scanning line by line, and the scan is like a matrix, because deep learning and convolutional neural networks rely on Recognize the pattern, but Livox’s point cloud is petal-shaped and looks different every time it is scanned. Machine learning is played around and requires many times the amount of data to learn.”
DJI is not unfamiliar with lidar, and it has deep enough optoelectronic technology and hardware manufacturing accumulation in UAV and camera products. If you can give full play to your strengths, you can cut away a piece of cake in this multi-billion-dollar market. But why did DJI choose such a “non-mainstream” solution?
In the past five years, the “immaturity” of lidar has been one of the core factors restricting the implementation of autonomous driving. The 40-line or 64-line mechanical lidar on the market not only costs about 100,000 to 250,000 yuan per unit, but it is also difficult to pass the “service life” and other vehicle regulations. Although the pure solid-state lidar has superior performance, the technology is not Mature and unable to be applied on a large scale; MEMS-based hybrid solid-state solutions are not yet capable of energy production.
With 15 years of experience in drones, DJI has a clearer understanding of “mass production”, “stability” and “price” than other lidar manufacturers. Back then, a large number of UAV competitors went bankrupt and went bankrupt due to “non-energy production”, “aircraft bombing”, and “cost cannot be held down”.
Moreover, DJI, which has a “learning from the past,” pays special attention to “patent safety”, and the unique solution can avoid patent disputes to a large extent. Before this, He said Technology and Velodyne, the “originator of lidar”, spent hundreds of millions to solve patent problems. DJI has also fought with Dautong Intelligent on the issue of drone patents many times and cost a lot of money.
Therefore, Livox made a risky decision: adopting a “rotating prism” hybrid solid-state solution, using “non-repetitive scanning technology”, and the point cloud presents a petal shape, which is completely different from the “repetitive scanning” solution of all other players. If customers want to use Livox’s lidar, they must redevelop the algorithm and adapt it. But this is a new cost to customers, and it also makes the early promotion of Livox very difficult.
Now, Livox has to improve the defects. 36 Krypton checked its latest product package and found that they began to provide some customers with firmware compatible with repeated scanning technology, a complete set of open source algorithms, and even higher-level algorithm support.
Although “complaints are full of complaints”, many industry insiders said that Livox’s lidar can basically meet the three conditions of “car passing regulations”, “large-scale stable delivery”, and “low cost” at the same time this year, without obvious shortcomings. If the OEM wants to increase the “selling point” this year, and put the lidar on the car, Livox must be one of the core choices. Xiaopeng P5 plans to carry two Livox car-grade lidar HAPs, which will be mass-produced and delivered by the end of the year.
“In the early stage of finalizing and selecting parameters, Livox will give a bunch of risk items. Later, I met another manufacturer, and there are almost no risk items. At first glance, it is very happy, but it is not true. The whole industry has not solved the problem, How did it solve it? Instead of telling me about the risks after getting in the car for a test, it is better to tell me at the beginning.” A source believes that the Livox team is “very real.”
Although Livox already has several car company customers, an automobile industry person who has been in contact with Livox told 36 krypton that the models equipped with Livox lidar have not been delivered on a large scale, and it is difficult to say how strong the “reliability” is. But in general, the later the MEMS-based hybrid solid-state lidar is shipped in batches, the more opportunities Livox has.
Should DJI build a complete car?
After going around, DJI’s shareholders told 36Kr that they are now suggesting that DJI build a car. This year, companies that have nothing to do with cars, Xiaomi, 360, OPPO… have all announced “cars.” But Wang Tao is still very hesitant about “building a car”, “feeling that the time is not yet ripe.”
In fact, Wang Tao wanted to build a complete car as early as 2016. At that time, Weilai had only been established for two years, and it was not too late to enter the game. Moreover, unlike Internet people such as Li Bin, He Xiaopeng, and Li Xiang, Wang Tao is the real “hardware tycoon”-in 2016, DJI has independently developed The battery management system and power drive system of the drone were built, and the chip team was built, and many years of engineering experience were accumulated-but Wang Tao did not directly build a car, but decided to invest in a sports car company first.
“New car companies are all raised by capital, and they are particularly accustomed to financing, even embracing,” a person who knows Wang Tao told 36 krypton, “but Wang Tao is different. He first used his family’s money and A friend (later became a shareholder of DJI) invested money, made two airplanes, sold two airplanes, made money, and then made the next airplane, otherwise there would be no material cost. He didn’t like capital operation, he was not a gambler, and never What listing plan is announced.”
For newcomers to the car, building a sports car is a good way to “test the waters”, because it does not need to spend a lot of money and build a production line. Tesla and NIO’s roads to build cars also started with the Roadster and EP9 sports cars. After accumulating experience in building cars and raising their brand image, they launched mass-produced cars.
After looking around, Wang Tao fell in love with Koenigsegg, a Swedish supercar company. This brand is not well-known in China and is hard to see on the road, but the average price is as high as one or two million euros. Its performance is no less than that of Ferrari, Porsche, Bugatti and other top supercars. Its hybrid supercar Regera has used it. In September 2019, it set the latest world record of 0–400km/h–0 in only 31.49 seconds.
One point that DJI is very attractive to Koenigsegg is: DJI can use its carbon fiber factory in Shenzhen to help Koenigsegg build car bodies, make parts, and provide a series of electronic parts and car-machine software to help it reduce costs, Increasing production capacity-Koenigsegg’s production capacity was only a dozen units a year at the time, which was expensive due to manual construction. As a result, DJI invested 20 million US dollars at the end of 2017 and obtained 14.89% of Koenigsegg’s shares. Wu Di, then the chief scientist of DJI, began to serve as a director in Koenigsegg.
This idea is feasible, and Li Shufu also made the same calculation: In 2017, Geely took a fancy to Ferrari and Porsche’s top British sports car company Lotus (Lotus), and injected 51 million pounds (about 460 million yuan). But the difference is that Geely is a holding company rather than a stake in Lotus; moreover, after 3 years, there have been clear results-last year, it has begun mass production and delivery of Lotus’s first pure electric supercar Evija, which is limited to 130 units worldwide. It was quickly sold out, and its pure electric SUV Lambda is also planned to be put into production next year.
But DJI did not have the luck and step by step of Geely, and Koenigsegg did not succeed in helping DJI build cars.
36Kr learned that the actual cooperation between the two parties did not produce enough “fetters”. Although the new energy vehicle manufacturing plan was formulated during the investment, frictions occurred during the cooperation process and the cooperation was not promoted as planned.
At the turn of the spring and summer of 2018, many employees of DJI were participating in Koenigsegg’s R&D project to develop multimedia functions for the car, but the cooperation project was cut off before the work was finished. The grassroots employees still don’t know the reason for the quarrel between the two sides. It was only half a year later from an announcement from the Hong Kong Stock Exchange that DJI “maybe” no longer a shareholder of Koenigsegg — —
In January 2019, Koenigsegg announced that it had received a 150 million euro investment from NEVS, sold 20% of its equity, and formed a joint venture with NEVS to produce new energy vehicles. The controlling shareholder of NEVS is Evergrande Health, a subsidiary of Evergrande Group. The premise of this investment is that DJI will not exercise the right of first refusal or exit completely. 36Kr asked DJI officials about the current shares, and the other party said that “there is no more information to provide.”
But what is certain is that after Evergrande’s founder Xu Jiayin made a high-profile announcement of building a car in November 2019, in March of the following year, Evergrande and Koenigsegg released their first cooperative product-the new energy supercar Gemera. Limited to 300 units worldwide, with a starting price of US$1.7 million, it is the world’s first Mega-GT model. The source told 36 krypton that this car was originally part of DJI and Koenigsegg’s “co-build car plan”.
DJI, which was one step ahead and rushed to the “strategic coordination”, ultimately failed to build a car through this investment.
Then, the time comes to 2021. The new car industry has experienced the ups and downs of NIO’s “nine deaths”, from the trough to the highlight, and the “build your own car” program has been placed on the desk of many big bosses.
Investors suggest that the logic of DJI’s car manufacturing is: DJI is good at making To C products and is a well-known brand, especially in the past 8 years in the European and American markets. If the smart electric car with the “DJI” logo goes overseas, it will be early Sales are likely to surpass “Wei Xiaoli” which makes foreigners unfamiliar.
However, people familiar with the matter told 36Kr that DJI is not a listed company and believes that it does not need to “build cars” to increase its market value. Wang Tao’s idea of the brand is also very different from others — Xiaomi is good at operating “Mifen”, and NIO has achieved the BBA price range for the brand within a few years. Wang Tao feels that the core value of DJI is not the brand. It is technology. He will not use this brand to expand categories everywhere, but rather expand his business from the derivation of technology.
Inside, one of the more pressing motives for “going to build cars” is that most of the cooperation between smart driving suppliers and car companies is just a transitional and mutual use relationship. Many car companies may eventually abandon their suppliers, self-developed autonomous driving algorithms, and master algorithm iterations and OTA upgrades. At that time, DJI may be forced to build a car, forming its own closed loop.
However, DJI’s past accumulation of brands, manufacturing, capital and sales channels is not enough to give it an absolute advantage in “making cars”:
From the perspective of money, from 2017 to 2020, DJI’s annual net profit will be about 4 billion-6 billion. Moreover, according to public information, in the bidding financing in 2018, DJI received about 1 billion U.S. dollars in the financing of Class B common stocks and 1.5 billion U.S. dollars in Class D common stocks (Class D shares are equivalent to “interest-free debt”, and three years have passed. period).
DJI is undoubtedly still a very rich company that can spend 10 billion to 20 billion yuan to build cars on its own, but if you consider that everyone is fully funded-Xiaomi claims that it has prepared 10 billion US dollars to build cars, Weilai ideal Xiaopeng All of them are issuing additional financing in the capital market to prepare for self-research of technology and the development of new models. DJI may have to break its habit of not loving financing to ensure the safety of cash flow.
From the perspective of sales channels, DJI now focuses on online sales, and has almost no offline direct stores. Last year, it transferred a total of 4 direct-operated flagship stores in China (Shenzhen, Shanghai, Beijing, Hangzhou) to agents. The partners of the company have saved the labor, rent and management costs of offline stores. However, if you want to sell cars in the future, offline store resources need to be rebuilt. After all, cars must be tested and driven in the store to realize the purchase of large items. Car companies are vying for a large-scale store on the first floor of the scarce offline supermarkets.
In terms of employment, DJI Automotive described their ideal candidates in the official recruitment post in May this year:
“He can tell clearly what he wants, and he has a story about why he wants to make things happen that is a piece of his real-life experience, maybe he set up a model aircraft himself, he made some small robots, he touched the capacitor and the screw with his hand. He remoted the shell and saw it fell on the ground and exploded, and it flew successfully. He has a very skin-conscious and very real-life experience of why he likes this thing. . It is not a sentence or a word, but his earthy breath, touch, taste and dense life at that time. He has a strong desire, an impulse to bring beautiful things to the real world. I think this may be an essence of R&D: think about some beautiful things that don’t exist in real life, and then bring them to this world.”
Between the lines in this passage is the shadow of Wang Tao, which is also the origin of DJI’s “engineer culture”. Obviously, he is using the standard of drones to recruit and select talents for the automotive team.
Many people who have been in contact with Wang Tao commented on 36Kr: he is very logical, extremely smart, diligent, strong in execution, very pragmatic, and pursues the ultimate. In an interview with 36 krypton, a drone colleague of DJI frequently expressed dissatisfaction with Wang Tao, but in the end, he unconsciously gave a thumbs up: “His product is really good!” But under the reputation, Some people say that he is not good at communication, is a bit too powerful, and the board of directors can’t control him.
All the advantages and disadvantages of Wang Tao’s talent and character helped him create this billion-dollar company. “This road cannot be copied at all. No one has successfully copied it for so many years. It is also very difficult for him to get out of this circle.” A source who knows DJI said with emotion.
The writer Plutarch of the Roman Empire once put forward a paradox about the change of identity, “The Ship of Theseus”, describing a ship that could sail for hundreds of years at sea, and attributed it to uninterrupted maintenance. And replacement parts. Whenever a plank rots, replace a new plank until all planks are replaced. Is this ship still the same?
For DJI, descending from the sky, breaking into the dense forest of cars, and fighting personally with the richest companies and smartest entrepreneurs in the world, to survive, it may have to change its nature or even replace itself. “Every wooden board.”
