China's TMT industry seems to go through a round of painful reflection every few years, and this time the new topic is whether AI can save China's software industry.
Adobe on the other side of the ocean uses a subscription system to pay for professional video editing users who cannot do without itself; Shopify uses a rent-collecting model to provide e-commerce sellers with the technology and templates to build online stores. A group of SaaS companies have led the traffic from Wall Street to the Greater Bay Area, and it has also made Chinese software companies smell the breath of "if you don't switch, you will die". Traditional software is similar to selling a house at one time, and building the next one after handing over the building; while the SaaS model is equivalent to long-term rent collection. Due to the high retention rate as the performance foundation and the non-linear profit growth after the break-even point, the stock price of the SaaS giant in the US stock market has a better match with its income, and it has a higher tolerance for early losses. The Chinese SaaS company whose valuation started, but in exchange for an article in 2023 "Investment Institutions Escape SaaS, Giant Companies Are Blowing Cold".
Chinese business owners may never have imagined that the same SaaS model will face all kinds of weird "customization" needs from all walks of life in China, which makes Chinese SaaS companies with "standardization" as the cornerstone of their business models start from research "Code code" has become "code head". Many entrepreneurs at that time, when customers kept asking for residual value, let go of their complacency at Zhongguancun Station of the subway and chose the oil and sauce separation in Nanluoguxiang.
The first batch of SaaS entrepreneurs thought about using PaaS + customized service outsourcing to solve the problem, but only a few office software came out; they also thought about turning IaaS into the cloud to save costs, but Almost no one thought that the MaaS theory, which originated in 2012 in AI technology, would become a life-saving straw to solve the persistent problems of the software industry ten years later.
Choice, perhaps really is more important than hard work. At that time, Zhang Shaofeng, who had just founded @百荣云-W（6608.HK）, might be among those entrepreneurs who were the first to understand the MaaS model.
The customization needs that SaaS companies have to face are joint-stock banks that need to improve the risk preference model to meet the financing needs of enterprises with insufficient collateral; it is the catering industry that needs to adjust the eight major cuisines with different spiciness and menu matching resulting in different full discounts; Inventory replenishment in peak seasons, big sales in off-seasons, transition to a retail industry where clothes are bought out of season and villas rely on the sea. SaaS companies are facing increasing labor costs to meet the actual needs of customers, but on the other hand, they are asked by the soul of the secondary market: "You are a company that relies on technology, why is the sales expense ratio so high?"
Since its inception, Bairong Cloud has been using AI to solve problems; SaaS is a business model for other companies, but for Bairong Cloud, it can only be a delivery method. It is possible that Bairong Cloud started to try the MaaS model because it had an idea to make as many data and models as possible and let customers call it to meet their own needs.
The word MaaS first appeared in 2012. To put it simply, MaaS is divided into three layers. The basic layer covers the most basic information of users; the middle core layer describes important user characteristics, such as user preferences, goals, asset volume, Annual income, social relations, etc.; the extension layer contains personalized knowledge with user characteristics, as well as personalized model networks (such as what yield products customers like). Together, these three layers form a MaaS platform. This platform can be used as an independent service platform, between the PaaS layer and the SaaS layer.
So how does Bairong Cloud use MaaS to solve the shortage of SaaS? For example, for any two different end users, there is no obvious difference in the SaaS services they get. For example, institutional customers all need customer portrait ratings, but the two customers have different weight requirements for user information. For example, some customers It is even more necessary to analyze the annual income of users, but some have to start with e-commerce transaction information. Customers instinctively believe that Bairong Cloud, a company that uses SaaS as its delivery method, is here to help itself solve all individual needs. So when customers have customized needs, SaaS companies have to develop on a low-code platform for half a day. But if Bairong Cloud has a MaaS platform that can provide different information for different users, then what needs do users have, wouldn’t it be enough to directly call Bairong Cloud’s model?
For example, Bairong Cloud's client is a commercial organization. When potential users of this organization apply for product services, the commercial organization will send a model call request to Bairong Cloud Creation, and Bairong Cloud will then use the model to evaluate the user. Just said However, MaaS is a cloud platform with many functional modules, so a complete user analysis can be obtained, and scoring results can be returned to the organization, and the end user will get the matching product from the commercial organization. To put it bluntly, the needs of customers are based on the existing amount of information and product cases. As long as MaaS covers enough information and solutions, institutional customers can directly call the parts they want.
Therefore, MaaS is equivalent to using the model call method to query the portrait of your potential users, and what kind of products and services will be matched with a high probability. Bairong Cloud is also conducting user evaluation for institutional customers on the basis of providing MaaS platform calls. That is, the underlying technology of MaaS - decision-making AI.
But the institutional customers at that time may not even have a complete understanding of SaaS, so it is unlikely that they will know what MaaS is. After all, the capital market and the enthusiasm for Internet + did not see the AI in 2023 at all. Become the only Sky Eye of the main line. As Zhang Shaofeng also said at the 2023 mid-year report performance conference, the underlying technology of Bairong Cloud is always AI, and SaaS is only a delivery method, not a business model.
No one knows about AI in 9 years, but one sentence of ChatGPT is known all over the world. In fact, from the first time commercial organizations cooperated with Bairong Cloud, they used the MaaS platform.
According to the recently released Bairong Cloud 2023 mid-year report, the company achieved a total revenue of 1.243 billion yuan in the first half of the year, a year-on-year increase of 32%; net profit reached 207 million yuan, a year-on-year increase of 121%. Bairong Yunchuang’s MaaS business revenue reached 431 million yuan, a year-on-year increase of 24%, and the core customer retention rate remained at a high level of 98%; the number of core customers was 146, a year-on-year increase of 14; the average income of core customers was 2.43 million yuan, a year-on-year increase 11%.
Can AI change China's software industry? I believe it is possible, because users can use natural language to interact with AI, which directly lowers the threshold for using B-end AI products. With AI after joining, non-professional video editor creators who used to be short video platforms may be able to use PR to automatically cut videos without having to learn how to add black frames and still frames; non-professional photographers can also use AI for themselves Send instructions and use the previous B-side PS products to help me repair pictures in Moments. Bairong Cloud, which is also rooted in the B-side, will its AI professional functions be far away from the C-side users?
So if a company with AI as the underlying technology can finally use MaaS as its business caliber and open poetry and Yuanfang, how many times will it be worth PS?
Source: Hong Kong Stock Research Institute