Low code, can not grab the job of programmers

Image source @ Visual China

The | of the text burns deeply, and the author | Wanqi, editor | Li Qiuhan

Recently, the "low code" that claims to be potentially losing programmers has attracted renewed attention.

According to media reports, the low-code manufacturer HeipaYun was acquired by ByteDance's Feishu team and will stop serving at the end of May this year, and its founder Chen Jinzhou will join ByteDance. The acquisition is considered by some industry insiders to be a reshuffle of low-code/zero-code platforms.

So-called low-code development, that is, no coding or only a small amount of code is required to quickly build applications. For enterprises, this can reduce the threshold of IT technology, so that more front-line business personnel can participate in or even lead the development of projects.

Internet giants have long targeted the low-code market. In December 2020, Alibaba Cloud connected the low-code development platform "Yideng" to DingTalk; in 2021, Tencent Cloud launched the "micro-tie" low-code development platform; DingTalk low-code platform "DingTalk" united with low-code vendors, and also established a low-code alliance.

Behind the sought after appearance, there are many problems to be solved with low code.

The low-code development that the outside world thinks only requires ordinary operators to drag, pull and drag in the background to create and complete the corresponding system operations, but in fact, in the application of technology, the current low-code operations still need to be completed by programmers with certain technologies.

"The low threshold of a low-code scheme and the ability to be fully customized, which are in conflict, are the most prominent contradictions in this field." At present, most low-code schemes mainly rely on a large number of templates to reduce the difficulty of users," said Liu Xing, a senior IT technician. This has made some companies hesitant to use low-code platforms.

In terms of commercialization, low-code vendors have also not gone well. Whether it is for medium to large enterprise customers or small business customers, a clear business path has not yet been explored. The black payun established in 2019 was acquired, and some insiders believe that to a certain extent, its performance is not ideal, according to media reports, some insiders said that the manufacturer currently has only 200 paying customers.

Low code, can it really be "everyone is a developer"? Is the market really in demand or a bubble?

Can you do "everyone is a low-code developer"?

In the development of low code, there is a popular saying: everyone is a low code developer.

Low-code platforms can be used in a variety of industries, helping users reduce work cycles and accelerate business progress. "Different low-code products can solve problems in different scenarios", an IT staff member introduced, for example, with the help of low-code platforms, merchants can complete the collection of voting information online, without the intervention of professionally trained technicians.

At the 2022 DingTalk conference in March this year, Ye Zhouquan, founder of DingTalk Yitao, said that DingTalk Yitao supports three new models of business innovation in the organization, one of which is "everyone is a low-code developer".

But can it really be done?

In 2019, Liu Xing's team developed a low-code platform relying on the WeChat Mini Program system to try to make "the whole people" participate. However, in the process of testing, they found that this still tests the user's "logical completeness" and "thinking ability", and the learning cost is not low. After a year of construction, the project was finally not officially launched, and it was announced by the company that it was aborted.

"At the C-end, it is currently impossible to achieve universal access to the whole people," Liu Xing said. With the ability to build low code, it is impossible to meet the personality, special customization, when the need to use a low-code platform to complete, complex to a certain extent, will make users daunted.

The reason, he argues, is that low code is eventually translated into code, and the programming language of code is contrary to the intuitive thinking of ordinary users, and it is very difficult to eliminate this difference.

In the wider enterprise, low code still requires users to spend learning costs and communication costs.

A business using a low-code platform starts with a time cost. "In the specific business process of the enterprise, if it involves special needs, when the technical staff of the enterprise puts forward the requirements and feedback to the low-code supplier, the supplier redevelops and then delivers to the enterprise, the whole communication takes time," Liu Xing said.

According to the feedback from past customers, a technician from a medium to large enterprise enterprise, who is exposed to the fact that it takes about a month for a technician from a medium-sized and large enterprise to contact their low-code platform from unfamiliar to completely usable.

Even in DingTalk, Ye Zhouquan also mentioned that in DingTalk, 60%-70% of low-code developers are front-line business personnel. This also means that there is still more than 30% of the demand, which needs to be done by non-front-line business personnel.

To solve these problems, low-code vendors need continuous optimization, and it's not yet fully mature.

Liu Xing took an e-commerce company that needs to do screening in an Excel form as an example, he introduced that each salesperson may have a variety of needs, customized and personalized needs for ordinary users, and has challenges for low codes. "Some vendors provide generic templates for specific use cases," he says, not really "everyone is a developer."

At the same time, the threshold for use still exists, "the large number of templates provided, users also have the cost of understanding these templates", Liu Xing mentioned.

Another industry problem is that "the low-code track, whether it is a new startup supplier or a supplier that has been raised, the chaos is that the product is not an open source framework developed by itself", Zhou Yang said, in the after-sales service, if the supplier uses an original open source framework, a problem in a certain link can be quickly found. However, if it is not an original open source framework, the logic of the open source framework is not clear, and tens of millions of lines of code are difficult to find, which also affects the customer's experience.

In general, low-code platforms, as commercial products, should be easy for users to get started, and reduce learning costs and time costs. But as a programming language, it still has a threshold of understanding and the difficulty of learning. "There are contradictions between the two that are difficult to reconcile," Zhou Yang said.

It is difficult to make money, and it is still impossible to win the big factory

In the early days, low code was only used in a small B-end market. Around 2016, with the strong pursuit of capital, many BPM (business process management) and form and form service providers transformed, entering this track and intensifying the competitive landscape.

How to quickly recover costs is a problem facing suppliers.

Zhou Yang introduced that the low-code platform can meet the risk control of medium and large enterprises for information deployment, business process management, process modeling, etc., and is also suitable for growing enterprises and start-ups. "Generally speaking, companies support collaborative work by purchasing third-party plug-in applications. Currently, suppliers are charged on an annual basis or per person," he said.

In 2020, low codes begin to evolve, which can be divided into two categories, one is low code for professionals and the other is zero code for business people. For suppliers, there are two markets for choice, B-end and C-end.

Zhou Yang, who has tested C-end low-code platforms, said that in the C-end market, users have their own ideas, and it is difficult to have uniformity. He and the team's product manager, after the project failed, summed up the experience and said that "it is difficult to achieve universality".

In the commercialization of the C-end market, it is difficult for suppliers to achieve profitability purely by selling low-code platforms. "In the case of low user awareness of the product, it is difficult for users to pay directly, which will also limit the user's motivation, and it is more suitable to build influence by doing the C-end," Zhou Yang said. They were working on a platform that used low-code development for public education, and originally envisioned recovering costs by selling textbooks and so on on the platform, but in the end they failed to run through this model.

For the B-end market, if enterprises have requirements for low-code platforms, they should consider whether they really reduce costs for them and whether they can complete the personalized needs of enterprises. Liu Xing said that low code does not rely on the system and cannot play a role alone. This leads to the nature of selling low codes, which is the sales system. Low-code vendors have a system and then customize the requirements according to different enterprises. When suppliers grab customers, they are actually competing for market share.

In this track, low-code suppliers are not competitive with big factories.

Zhou Yang introduced, the market mature enterprise-level collaborative office platform DingTalk, enterprise WeChat, etc., also launched their own low-code applications, if it is a small and medium-sized enterprise, the demand is not large, the cost of purchasing a low-code system alone is not low, directly use DingTalk, enterprise WeChat, you can meet the demand. At present, the enterprise-level collaborative office platform developed by the Internet factory has occupied a certain market volume. Some suppliers will also rely on DingTalk, Feishu, and enterprise WeChat systems to sell low-code services.

For larger medium and large enterprises, Zhou Yang introduced that suppliers often adopt the "rental cooperation" model, and then sell product authorization to customers, collect service fees, etc. Taking the products launched by his company as an example, the service benchmarked with the OA system can be rented in units of 30 people, and a rental fee of 2600 yuan + can be charged.

But the challenge for suppliers is that their systems are not versatile and costly. "Only when a set of products is sold to as many customers as possible, the marginal cost will fall. When developing a system for each customer, there will be a lot of communication and adjustment," Zhou Xing said.

In the eyes of some industry insiders, ByteDance's acquisition of low-code manufacturers Black Payun means that the low-code market has opened a game of "big fish eating small fish". The entry of large Internet manufacturers to stir up the market pattern, other suppliers want to break through, it is not easy.

What are the prospects for the future?

What is the future of low code?

According to the "Ecological Aggregation - Insight Report on the Ecological Development of China's Low-Code Industry" released by iResearch in 2022, from the current industry penetration status, the penetration rate of labor-intensive industries such as manufacturing, pan-Internet, and education is relatively high, and the penetration rate of low-code applications in the manufacturing industry is expected to be about 20%.

According to the situation that Deep Burn understands, from the perspective of micro demand, "the business needs of governments and enterprises and central enterprises are extensive, and there are many places that need to be built and applied." If they are self-developed, they will encounter many problems in development and the cost is also high. When you buy low-code products, you only need one or two technicians to build them," Zhou Yang said.

Positioning the suppliers of B-end enterprises, according to the analysis of T Research Report, in 2020, the return on investment is considerable, and the return on investment in the overall track is positive, because customers are mainly concentrated in medium and large enterprises, have demand for low-code platforms, and have purchasing power.

According to Forrester Consulting, the market for low-code development platforms will increase from $3.8 billion in 2017 to $21.2 billion in 2022, with promising prospects.

However, at present, this track still has problems to be solved in many aspects such as technology, market, and capital.

Technically, low code doesn't make it possible for every average operator to participate in the creation. "At present, the maturity of low code is still at the level of need for technicians to communicate between front-line salesmen and developers." An industry insider said.

Applied to the market, for enterprises, "whether enterprises are willing to spend energy on training employees, or whether employees are willing to learn, this is the reason why enterprises are hesitant to use low-code platforms," Liu Xing said. This is also the reason why it is difficult for suppliers to develop small and medium-sized enterprises.

In terms of capital, the domestic low-code track, in 2015, there were a large number of low-code suppliers, and since 2018, capital has followed up, and the number of investment and financing has increased. However, according to T Research's "2021 China Low-Code/Zero-Code Panoramic Industry Research Report", in terms of financing level, low-code suppliers are mostly in the angel round and A+ round, no more than 50 financing brands, and a large number of waist and small and medium-sized suppliers are still in the incubation period.

At present, a large number of waist enterprises and small enterprises have insufficient understanding of low codes, and the degree of informatization of domestic customers is still in its early stages.

According to T Research, in the low-code track, it is still dominated by medium and large enterprise customers with thickER IT technology. In the enterprise market, more than 71.5% of enterprises have not heard of low-code related technologies, and their awareness of low codes is still at a low level. Less than 10% of companies that have knowledge, try, or have already used it.

Zhou Yang told Shen Yan that at present, his company set the goal of developing small and medium-sized enterprise customers last year, but after contacting them, these companies do not know much about "what is low code".

This makes the use of low code in the market, and it is still impossible to clearly see the way forward, and the day when programmers can be laid off collectively is far from coming.

*At the request of the interviewees, Liu Xing and Zhou Yang are pseudonyms. (This article was first published on the Titanium Media APP)