Image source @ Visual China
Wen | parity
For most companies in emerging industries, being held high and falling down seems to be a fate.
In the AI companies that have been listed, Geling Deep Pupil was once debated by the big guys Xu Xiaoping and Shen Nanpeng about "how many hundreds of billions of dollars in the end", but now its market value is only 4 billion, and the unit is RMB.
Yuncong Technology, one of the four little dragons of AI, was finally approved by the CSRC this month to apply for registration on the Science and Technology Innovation Board after two inquiries.
Behind this is the tortuous development of the AI industry in recent years. In 2016, Alpha Go swept the Go world champion Lee Sedol, shocking the entire technology community. The public began to have unlimited reverie about the future application scenarios of AI, and the media poured in to publicize it, and AI became a momentum myth.
But after three or four years, people gradually found that AI is not as powerful as imagined, and the changes to life are not too much. The negatives of AI companies that used to have unlimited scenery have also begun to appear frequently in the news, and the road to listing has been quite difficult.
Among many AI companies, the cloud from technology is undoubtedly lucky, at least it has survived to this moment of listing financing, but cloud from technology is also unfortunate, in the PAST few years of the AI industry, cloud from technology has accumulated a lot of problems, so that in the end can only "bleed listing".
The rise, climax, and then fall of the AI industry is also the epitome of the development of the cloud from science and technology. Through this article, we try to analyze cloud from technology from the following three questions:
1. Opening scenery financing billions, why does Cloud From Technology still insist on "bleeding" listing?
2. Cloudcong Technology, which is mired in losses, what is its operating conditions and business?
3. If you want to reverse the decline, in addition to listing blood, what else can Cloud Technology do?
Perfect start, it is difficult to solve the huge losses in a row
As one of the four ai dragons, the cloud has a rather scenic start from technology.
Much of this depends on the experienced founder Zhou Xi. In 2006, Zhou Xi studied for a doctorate at the University of Illinois at Urbana-Champaign (hereinafter referred to as "UIUC") under the tutelage of Professor Huang Xutao, the "father of computer vision". During his time in school, Zhou Xi showed his talent in the direction of AI vision to the fullest.
He has led the team to participate in a number of international competitions on behalf of UIUC, defeated famous research institutions such as MIT, the University of Tokyo, and IBM, won the world championship of six intelligent recognition competitions in succession, and published dozens of papers, and the impact factor of the paper was amazingly high.
Zhou Xi, who graduated with a doctorate, naturally did not worry about high salary offers, but Yuan Jiahu, a big man of the Chinese Academy of Sciences, noticed the top AI rookie of the year and invited him to return to China three times to invite him to return to China for development. So Zhou Xi finally chose to return to China in 2011 and co-founded the Chongqing Research Institute of the Chinese Academy of Sciences to engage in research on image vision.
This has also become the foreshadowing of Zhou Xi's future establishment of CloudCong Technology, and an important endorsement of Yuncong Technology in the financing process. In the five years of the Chinese Academy of Sciences, Zhou Xi discovered the powerful charm of AI vision and its vast market. In 2016, Zhou Xi decided to give up the preferential treatment of the Chinese Academy of Sciences and establish Yuncong Technology.
There is a doctor who graduated from the United States and was the founder of the local branch research institute of the Chinese Academy of Sciences, and the financing process of cloud from science and technology is naturally quite smooth. In the angel round alone, Cloud received 60 million yuan of financing from Technology, and at the end of the same year, Cloud received another 200 million yuan increase from Technology.
In the following years, as the concept of AI prevailed in China, a large number of technicians devoted themselves to algorithm research, and companies with AI-related concepts were even more popular. In 2017, a large number of AI vision companies received large financing, such as Megvii Technology received a C series investment of 36 million US dollars.
However, among many enterprises, Yuncong Technology founded by Zhou Xi is still one of the most optimistic AI companies, and investment institutions with state-owned backgrounds such as Yuanhe Origin and Guangzhou Fund are very optimistic about the market potential of Yuncong Technology, directly investing 500 million yuan in the B round.
During this period, the "AI Four Little Dragons" pattern has also been established, in addition to SenseTime, Megvii, Yitu three, cloud from the technology is not surprisingly ranked among them, the founder halo blessing, the financing amount has gone all the way up, the perfect start of cloud from science and technology is still continuing, and then the three rounds of financing, cloud from science and technology has obtained 1 billion yuan, 310 million yuan and 1.8 billion yuan of blessing.
In the five years since its establishment, Yun has raised billions of dollars from Technology Wind and Solar, and its valuation is close to 15 billion. Unicorns, the most promising enterprises and other awards, the cloud from science and technology to get soft, by the capital to the spotlight it is particularly dazzling, funds should not become an obstacle to the development of cloud from science and technology.
However, this time listing, the evaluation of the cloud from science and technology is mostly "bloody" and "bloody" listing evaluation. The cloud from the perfect start of science and technology, did not let it maintain the scenery of the past, you are one of the "four little dragons of AI", the cloud from technology or to the point of having to go public to raise funds.
On the one hand, with the development of the industry, AI companies are leaving the stage of "storytelling" and beginning to face the reality of industrial landing. On the other hand, the cloud faces fierce competition in the field of image recognition where technology is located, and whether it is the practical application of the landing or the process of competition with other enterprises, the cloud from technology can not avoid a large amount of burning money.
This also makes Cloud From Technology, although it has raised billions of dollars, it is still in a deep loss dilemma.
According to the latest prospectus submitted by Cloud From Technology, from 2019 to 2021, the main business income of Cloud From Technology was 780 million yuan, 751 million yuan and 1.07 billion yuan, respectively, while the losses in the three years were 640 million yuan, 813 million yuan and 632 million yuan, respectively. The amount of loss during the three-year period was as high as 2.085 billion yuan.
The financing brought by the beginning of the past scenery has been continuously consumed with development and competition, but at this stage, the hematopoietic ability of Cloud From Technology cannot completely offset the company's decline in losses, and the huge losses that Cloud from Technology is difficult to avoid are the most important reasons for its continued dedication to financing and listing.
Source: Cloud from Technology Prospectus
If you can't make ends meet, where is the money "pit"?
With consecutive years of huge losses, coupled with the inability of revenue scale and enterprise valuation to match, cloud applications for listing from technology had to go through two rounds of inquiries, and even the review was suspended for a time.
So, what exactly is the reason for the cloud's long-term losses from technology, and now it faces the dilemma of capital flow. The cloud can find some reasons in the technology's prospectus.
The ratio of R&D investment is higher than the normal level of technology companies. In 2019, 2020 and 2021, the company's R&D expenses were 454 million yuan, 578 million yuan and 534 million yuan respectively, and the cumulative R&D investment in the past three years totaled more than 1.566 billion yuan; the company's cumulative R&D investment in the past three years accounted for 59.39% of the cumulative operating income in the past three years.
Source: Cloud from Technology Prospectus
In the AI industry, this seems to be an unsolvable problem, cloud from science and technology as a typical high-tech algorithm enterprises, R & D investment has a key impact on the accuracy of the algorithm, in order to make their own algorithm accuracy and response time is not defeated by peers, related AI companies are constantly increasing research and development efforts, resulting in a serious internal volume situation in the industry.
At the same time, the competition for talents among peers is also constantly improving the salary level of algorithm technicians, which further aggravates the labor costs of enterprises.
The result of the inner volume has caused the four small tigers of AI, including cloud technology, to have no profit at present, and all of them have huge losses.
In addition to competition from the same industry, giants are also announcing their entry into AI. The field of AI seems to have become a battlefield for the crowd. Ali, Tencent, Baidu, Huawei and many other powerful manufacturers have also begun to layout, in the face of many well-funded competitors, the cloud from the technology, whether it is capital or traffic or talent reserves, is undoubtedly in a weak side, it is difficult to match the opponent.
In addition, camera hardware manufacturers are also expanding the AI market, such as Hikvision and Dahua shares, research and development investment Hikvision three years of R & D investment of more than 20 billion yuan, Dahua shares also exceeded 9.2 billion yuan.
Although the total scale of R&D of Cloud From Technology exceeds its normal level, it is not worth mentioning compared with domestic leading enterprises.
If technology is the driving development of enterprises, then the current state of cloud technology is that it has done its best, but found that its investment is less than a fraction of that of the head enterprise.
Parity mapping
Before there is the internal volume competition of peers, and then there are Internet giants and hardware manufacturers with high funds to enter, their joining will bring great challenges to the tight cloud from technology. If the cloud technology does not want to be surpassed or annexed by large factories, it can only continue to develop to deepen its own moat, which has caused high investment in research and development.
Lower gross margins became another reason why the cloud was mired in losses from technology. From the perspective of the company's main business income, in the past three years, the revenue of "human-machine collaborative operating system products" accounted for 23.48%, 31.50% and 12.72% of the main business income, and the gross profit margin was 89.30%, 75.86% and 73.99% respectively; the revenue of "artificial intelligence solution products" accounted for 76.52%, 68.50% and 87.28% of the main business income, and the gross profit margin was 23.43%, 28.19% and 31.34%, respectively.
The composition of the company's main business income by business type is as follows:
Source: Cloud from Technology Prospectus
Overall, from 2019 to 2021, the gross profit margin of the main business of cloud technology is 38. 89％、43． 21% and 36. 76%, the average gross profit margin of the main business in the past three years is 40%. Compared with the gross profit margin of Megvii Technology products of 59%, the gross profit margin of Yitu Technology products is 63%, and the gross profit margin of SenseTime Technology products is 70%, the profitability of Cloud Technology is far less than the industry average, and the gross profit margin level ranks first in the "AI Four Little Tigers".
Parity mapping
The reason for the low gross profit margin of the main business of the cloud from technology is that the gross profit margin of the company's product "artificial intelligence solution" is relatively low, with an average of only 27.65% in three years, while the product accounts for 77.43% of the company's main business revenue, which reduces the gross profit margin of the company's overall products.
"Artificial intelligence solution products" are single products, which require installation, commissioning or custom development, and the higher the degree of customization, the higher the cost of outsourcing services and labor costs, resulting in relatively low gross profit margins of products.
The continuous superposition of accounts receivable has brought high bad debts to the cloud from technology. From 2019 to 2021, the balance of cloud receivables from technology was 308 million yuan, 523 million yuan and 420 million yuan respectively, accounting for 38%, respectively, accounting for 38.5% of the current operating income. 19％、69． 36% and 39. 07％。 As of the end of 2021, the amount of cloud bad debt provisions from technology has reached 67.1548 million yuan.
Source: Cloud from Technology Prospectus
The main reason why the amount of accounts receivable of the company continues to be at a high level is that the downstream customers of Yuncong Technology are mainly large government and enterprise customers such as banks, airports, and public security. Because the payment of government and enterprise projects needs to go through strict acceptance, special final account audit and fund approval process, and can only pay the payment after the corresponding special funds are in place, the payment time of some projects is later than the payment time agreed in the contract, and the payment collection cycle is longer.
Cloud from the technology main customers Source: screenshot of the official website
The continuous rise in accounts receivable not only brought heavy bad debt pressure to Yuncong Technology, but also made the company's operating cash flow continue to deteriorate. During the reporting period, the cash flow expenditure generated by the cloud from the operating activities of science and technology was 506 million yuan, 462 million yuan and 547 million yuan respectively, and the net outflow of operating cash flow in the three years has exceeded 1.5 billion yuan.
Listing is not the only "solving" answer
The cloud from the technology is now facing difficulties that many companies have encountered in the early stages of development. But there are still many companies that can turn over in adversity, which also indicates that the future of cloud technology is still expected to "solve difficulties" in various ways.
In this process, listing financing is often only the first step, and the cloud needs to do a lot of work to reverse the decline from technology.
For example, make the product more differentiated. Statistics found that many solutions from the cloud technology are similar to other companies, mainly focusing on the security, financial and retail industries, and homogeneous competition has also led to a further intensification of the involution effect.
Parity mapping
Around 2018, a set of face recognition algorithms can sell tens of millions, and now it is only worth 400,000 yuan, and the threshold of image recognition algorithms is getting lower and lower, which is a sigh issued by a technology bull after leaving an AI startup.
SoftHong Technology, which also provides AI image algorithms, has achieved the differentiation of its own products. In 2022, it will not only create a profit of 250 million yuan, but also the net profit margin of the product will be as high as 36.8%, which may be what the capital market expects from AI technology.
SoftHong Technology has its own subdivision of research, which is to provide photography optimization solutions for customers in the fields of smart phones and other fields, which is used to improve consumers' photography functions for mobile phones. Now customers include Huawei, Samsung, OPPO, VIVO, Xiaomi, Sony, LG, TRANSSION and other well-known mobile phone manufacturers.
After occupying a segment that has not been homogenized, the company can have a high profit margin. With the funds, SoftHong Technology has begun to provide visual solutions for intelligent driving in emerging markets, forming a virtuous circle, rather than expanding the Red Sea market such as security, finance and retail industries.
In addition, you can also try to do a combination of hardware and software products. If AI algorithms can't take over a niche, they can try to develop a combination of hardware and software. Build your own industrial ecology to prevent it from being replaced by upstream and downstream manufacturers.
The core of the business model of the cloud from the current stage of technology is to sell its own "model" and "computing power", and the way to achieve profitability is to let customers pay when they use it. Although the AI model and computing power look bright and beautiful, the appearance has a high-tech coat. However, the business development model is quite constrained by Party A, and it can never escape its identity as "technology outsourcing".
In order to reduce costs, customers who purchase AI products often choose to develop their own models and algorithms in the later stage. Taking the field of computer vision as an example, as an algorithm buyer, Hikvision, Dahua and other camera hardware manufacturers have been promoting the visual recognition algorithm business to build their own upstream and downstream ecology.
Technology may be imitated and overtaken, but the integration of technology into the industry and products can highlight the value of technology, achieve stable commercialization, and establish a moat that is difficult for competitors to cross.
Make as many products as possible replicable. The big thing about the cloud from technology products is "artificial intelligence solutions." In the past three years, the revenue of artificial intelligence solutions accounted for 77% of the main business revenue, but the gross profit margin was only 28%. The main reason for the low gross margin is that the product cannot be "copied" to the next customer.
Unlike the "human-machine collaborative operating system solution", the "artificial intelligence solution" only needs authorization and simple framework modification to become a new product. Instead, it requires installation and commissioning or custom development, which leads to high service costs and labor costs.
If you spend a lot of human and financial resources, the hard-earned products cannot be used by the next customer. After opening the next customer, it is necessary to redesign, develop, test and install, so that the gross profit margin of the product is naturally not high, and it will eventually affect the overall revenue and profit.
On the basis of the current customers, through the above three initiatives, Cloud from technology can get higher returns on R&D investment, to obtain sustained and stable revenue, and improve the gross profit margin of products. Thus alleviating the financial pressure caused by huge R&D investment, and avoiding the problem of insufficient hematopoietic capacity caused by low gross profit margin of products.
With the implementation of the above initiatives, the improved cloud from technology can also expand into new areas of customers, thereby further shortening the receivables period.
This may be a more practical way to "solve the problem" of cloud conglomerate technology, which is deeply losing money, in addition to listing financing.
Write at the end
The well-known Gartner consulting company has proposed a technology maturity curve, which divides the maturity and evolution of new technologies and the time required to reach maturity into 5 stages: technology budding period, expected expansion period, bubble bursting trough period, steady climb recovery period and production maturity period.
Technology Maturity Curve Source: Network
It is not difficult to see that the cloud from science and technology is also in this cycle, in the embryonic stage of technology to see a vast AI market, capital continues to inject, all the way to the top, the expectation value is also more and more inflated; and the current stage of cloud from technology, is in the expectation of inflation period and bubble bursting trough period between.
At this stage, enterprises like Cloud From Technology are at a fork in the road to falling and taking off, and where the future is going depends on the choices made by Cloud from Technology.
After the "bloody listing", can the cloud return to the upward channel from technology? It may also be necessary to see whether it can solve the above problems and whether it can implement development programmes.
Resources:
"Feed the AI Four Little Dragons" photon planet
"AI to the vertical, the story of the "Four Little Dragons" is still good to tell? Luo Shao Business Review
"Cloud sprints from technology to IPO: Accelerate the landing of the industry, expected to be profitable in 2025" hot article life new horizons