From technological momentum to commercialization, there are still many lessons for AI companies to make up for IPO.

Editor’s note: This article is from the micro-channel public number “goodness Finance” (ID: IPOxscj), Author: good deeds Finance.

Recently, according to an announcement by the Beijing Securities Regulatory Bureau, Megvii Technology intends to go public in the form of CDR (public issuance of China Depository Receipts).

Megvii Technology stumbled all the way and finally came to the front of the capital market.

In 2017, Yin Qi, the founder of Megvii Technology, revealed that he aimed to go public, to the postponement of his IPO in Hong Kong in 2019, and now he is willing to enter the Science and Technology Edition for counselling. This difficult IPO road has been It’s been 3 years.

Now that the AI ​​trend has passed, the market has become more rational, and the result of the listing of Megvii Technology has cast a shadow over. Megvii Technology, which has repeatedly sought IPOs, has also ushered in the next turning point in its destiny. ,

01 Continuous losses, insufficient scene-based implementation,Kuang Is it easy to log in and create a version?

According to Tianyancha App financing information, Megvii’s latest round of financing occurred in May 2019, with investors including Alibaba and Bank of China Investment. From the perspective of the number of financing, from 2011 to 2019, Megvii Technology has received 7 rounds of financing.

Whether it is the needs of the capital market or the needs of the company’s own development, under the reality of continuous losses and AI technology scenarios that fall short of expectations, listing seems to be the only way out after multiple rounds of financing.

According to the previous prospectus published by Megvii, the annual revenue of Megvii Technology from 2016 to 2018 was 67 million yuan, 313.2 million yuan, and 1.4269 billion yuan respectively, showing a rapid growth trend, but at the same time, losses were also Is rising rapidly.

From 2016 to 2018, Megvii lost RMB 342.8 million, RMB 759 million and RMB 3.351 billion respectively. For the six months ended June 30, 2018 and June 30, 2019, losses were 729 million yuan and 5.2 million yuan, respectively, an increase of 613.3% year-on-year.

In addition to continuing losses, the company’sThe cash flow situation also does not seem to be so ideal. The data shows that the cash flow amount used in operating activities in 2017, 2018, and the first half of 2019 was -150 million yuan, -720 million yuan and -670 million yuan, respectively, which has been expanding. trend. This also reveals despise technology from the side

Megvii also seems to be facing the dilemma of insufficient growth momentum in terms of scene implementation.

In fact, the scenarios for AI visual scene applications are very limited. The more mature scenarios are mainly concentrated in the financial and security fields, followed by the smart city field. The former has insufficient market growth and limited growth space, although the latter has Good development prospects, but the industry competition environment is not optimistic.

According to the prospectus submitted last year, urban Internet of Things solutions accounted for the largest proportion of the three main revenues of Megvii. In 2019, the proportion of revenue has been reached, and the revenue is more inclined to G, and the giants get together In today’s smart city, this means that Megvii will inevitably face more powerful competitors.

In the G-end market, Internet giants are more likely to gain the trust of G-end users than startups, and they are also in a relatively advantageous position in market competition. This also means that in the G-end market, Megvii’s market competition pressure is actually very great.

Secondly, in the security field where the industrialization of AI technology is relatively mature, Megvii Technology also needs to directly compete with “indigenous giants” such as Hikvision and Dahua Technology. Compared with the former, the latter has a deeper industry accumulation. , Megvii Technology may not be so easy to compete for the market on a large scale.

Megvii has also said, “The industry we operate in is very competitive and we face competition in many major business areas. If we fail to stand out from our existing or future competitors, our business, Financial conditions and operating results may be materially and adversely affected.”

In the view of Xiangshan Finance, in the current industry and capital environment, there is no obvious gap in the technology of first-line computer vision companies. The real scene competition is actually the relationship and channel of the enterprise-level market. This means that the valuation logic of AI Vision will gradually change from “industry valuation” to “financial valuation”.

The so-called “industry valuation” means that in the capital market’s value judgment for an enterprise, the value base of an enterprise in the secondary market is determined based more on the market size, market share and technological advantages of the enterprise. More emphasis is placed on the company’s future growth space.

And “financial valuation” refers to the company’s current financial indicators, such as an annual (maybe three to five years) profit growth, cash flow status, and business revenue growth to determine the actual value of the company. In contrast, the former pays more attention to the growth of the company itself, while the latter can more truly reflect the actual value of the company.

The capital market’s return to rationality in the value judgment of AI companies also means that in the past, AI companies “tell good stories”.Getting a good valuation is a thing of the past. In other words, for Megvii, which desperately wants an IPO, even if it is difficult to go public, it may be difficult to have a relatively good market value performance.

02From technical momentum to commercialization, There are still many lessons for AI companies to make up for IPO

From the wave of AI financing in the past few years, to the “four little dragons” in the computer vision field of Shangtang, Megvii, Yuncong, and Yitu who have successively sought to go public, one problem cannot be avoided: commercial applications have become the industry leader. After the keynote, how to further meet the expectations of investors.

In fact, as an emerging technology-based industry, the short-term value of the industrialized application of AI technology has always been overestimated. The artificial intelligence industry has always created too much momentum, and in terms of landing, it has been There are indeed commercial scenes, but they are far from being anchored in the hearts of investors and the public.

Therefore, the IPO of AI companies requires a “commercialization advancement” from technology to industry.

In other words, if AI technology companies want to truly become a “high-value commodity” and go to the secondary trading market, there are still many lessons to make up.

For example, in the process of industrialization, how to deal with the relationship between AI technology product application and public privacy.

In August 2019, Megvii Technology was suspected of violating privacy for launching an AI product that monitors students in the classroom. At that time, Megvii responded that the online classroom behavior analysis was only a conceptual demonstration. Although the response from the despise side was fairly timely, repeatedly stepping on the “privacy landmine” can easily create the impression of “defeating and not paying attention to privacy protection”.

The “data AI” technology that has been gradually applied today has largely created huge conveniences for our lives, such as the convenience of offline facial recognition payment systems, and the effective AI facial recognition applications in the security field Improve the efficiency of security inspections, digital marketing based on user data portraits, etc.

But in such an industry environment, there is no effective “data black box” mechanism to ensure the absolute security of personal data. How to deal with the relationship between the application of AI vision technology and the protection of privacy? It may be a new topic,

In September last year, Kai-Fu Lee, CEO of Innovation Works mentioned in a speech that he helped Megvii technology companies find partners including Meitu and Ant Financial in the early stage, allowing them to get a lot of face data.

Li Kaifu’s remarks brought the issue of Megvii’s privacy protection to the forefront. Then the ants responded that “Privacy protection is the antThe lifeline and red line will never be violated at any time.

Megvii Technology also responded that it will not actively collect any personal information from end users” and “always attaches great importance to the protection of data privacy.” Although Megvii responded to outside questions in a timely manner, whether it can be completely eliminated The user’s privacy and security concerns are still unknown.

In the view of Xiangshan Finance, the repeated high attention of the outside world due to privacy protection may have a certain degree of impact on Megvii Technology’s ToG business. The privacy issue may also be a key node on the road to despise the IPO.

In fact, the application of AI technology at the G-side level may be more concerned with privacy protection than the B-side. For example, in the classroom surveillance turmoil that had been exposed before, although the contempt was explained afterwards, the influence of public opinion is difficult to completely eliminate, which means that even in the conventional security field, it may be difficult to open the education market.

“The impact of privacy issues on the business may also be further spread to the secondary market. For example, during the entry phase, the regulatory authorities may further investigate the possible problems of the company in terms of privacy, etc.” There are investment bankers Said to Xiangshan Finance.

From an industry perspective, Megvii Technology is also facing pressure from competitors even in the IPO.

In the CV (Computer Vision) Four Little Dragons, Yuncong Technology and Yitu Technology are also preparing for listing. In August last year, Yuncong Technology officially started IPO counseling, and received the first round of inquiries in December. Yitu Technology launched its IPO counselling in September last year, and the Shanghai Stock Exchange officially accepted its IPO application in November.

“In terms of valuation, companies that are listed first may have certain advantages.” The above-mentioned person said: “Generally speaking, in the absence of sufficient reference system, the valuation of the first AI vision companies may be relatively It is higher, and since the listed companies already have reliable references, their valuations may be relatively conservative.”

Wait for wind is not as good as wind. In order to convert investment value into market value and valuation, the capital market must see real results. This means that after the G-end business and the B-end business are done well, if the former value of the C-end market can be fully tapped and recognized by the C-end market, it is possible to realize an appreciation.

Specifically, we must adhere to the niche route, dig deep into the value of AI vision hardware products, and avoid direct competition with mainstream manufacturers. On the other hand, the bargaining power of C-side buyers is also relatively weak, unlike other mature products. , The market price has already stabilized, and it is easy to obtain higher profits.

In general, going public is not the end of technological innovation enterprises, but another starting point. For Megvii Technology, this IPO is an opportunity and a challenge. However, compared to the market value after the IPO, perhaps the integration of AI technology and market demand and the connection between the privacy and experience of the C-end mass population may be topics that require more in-depth research.