How to commercialize digital products? Why are some open source projects so good, but commercialization always fails? In the fourth issue of "Open Source Talk", we invited Wang Yezhuang (Tou Ge) and Li Qipeng (Jian Feng) to talk about how to commercialize open source projects.
-
Wang Yezhuang (Touge) , the host of "Touge Kanma", focuses on sharing content on topics such as technology, entrepreneurship and product innovation.
-
Li Qipeng (Jian Feng) is currently the CEO of vanus.ai. He has worked at Alibaba Cloud and Apache RocketMQ PMC, and has long been focused on AI infrastructure software and middleware.
Brother Tou: There is no direct relationship between technology, a piece of code, and a good product, only an indirect relationship. So you will find many people who do a good job in open source but do a poor job in commercialization. On the other hand, some people do very well in commercialization, but they don't do it in the community. Today's AI products, large model products, and open source products are actually digital products. So here comes the first question:
How to commercialize digital products? Is there any way in between?
Li Qipeng: Let me share the open source products first. Generally speaking, we will first open source the product on Github to attract some popularity. In this way, we can find the initial users (after all, open source is a good way of promotion), and secondly, we can quickly polish the product through open source, and with the help of developers, the product can be quickly iterated and matured quickly. The second step is commercialization. In China, if we do infrastructure, we usually sell it through licenses or get some big projects. We have served large clients before, such as banks, Saas companies, etc. In addition, there is another way to provide SaaS services. We will first host the product on the public cloud, so that users do not need to deploy or operate the product themselves when using it. It can be opened out of the box when needed. Just use it.
However, different projects have different shapes and things that need to be considered during commercialization. For example, our current users of vanus products are both domestic and foreign. But based on our experience, the difference is that in China, it is difficult for companies that provide software services to avoid privatized delivery methods. It is quite difficult to run a pure product company in China's current environment. I have worked for both Party A and Party B. My feeling is: in China’s software market, Party A is very powerful, and software users have a poor sense of boundaries. If they buy your product, they will He will hope that you can solve all problems for him, whether it is software problems or peripheral problems, you must solve them for him. Now even when our domestic users are recruiting, they have to ask us for suggestions. I even helped my client to conduct interviews (laughs~)
But having said that, this method still has its advantages. First of all, you will establish a strong connection with the customer. Because you have served for a long time, he will trust you and give you many things. It may be easier for you to get the order. For example, some Chinese government and enterprise customers have projects that last from one to three years. For your company, the gross profit may not be particularly high, but it will allow you to have continuous cash flow input, which will be beneficial to the development of the entire company. All have benefits.
Of course, the disadvantages of this approach are also obvious. For example, my current product serves dozens of customers. Each customer has a code branch, and each customer has a different code. This puts a lot of pressure on the maintenance and operation of our products. You will find that a software company will become bigger and bigger in the later stage. This does not mean that the product or business will become bigger and bigger, but that there will be more and more people, especially the implementation personnel and maintenance personnel. . In this way, the overall gross profit will be relatively low, and the products themselves will be relatively fragmented. In the end, there may be several products instead of one product. This can be regarded as an interesting Chinese characteristic.
From a market perspective, there are still relatively few pure product-based software companies in China, and more are project-based companies. But it is different overseas. From our experience, overseas customers have a very strong sense of boundaries and have high renewal rates, which is very helpful for you to spend time polishing your products. However, this method also has disadvantages, that is, he is unwilling to establish too many connections with you. We wanted to conduct a user interview before to ask about the product usage experience and see what other scenarios our product can meet, but we encountered difficulties. He may think, I can use this product and I paid for it, so why are you always looking for me?
Haha, in short, the difference in this market is very obvious. I personally think that these two forms are relatively typical, and each has its own pros and cons. If you want to start a startup project, you have to make a choice based on specific products, different team styles, and even the founder's style.
Brother Tou: I really feel what Qi Peng said just now. Let me share my experience. I started working relatively early. I started to be exposed to Java in 2001 and IOE architecture in 2004. There is a TV shopping in Shanghai called Oriental Shopping, which we did. Seven or eight years ago, there were no basic software manufacturers in China. A little earlier, fifteen years ago, if you were looking for a database, you could only think of Oracle. Third-party service companies from standard manufacturers are very popular. The reason why our Party A is so strong is because they were coaxed out by these people.
When I was shopping in Dongfang, we initially bought Oracle's original service. Oracle engineers came here to support us, and they were paid 10,000 yuan for 8 hours a day, plus overtime. Think about what level 10,000 a day was in 2004? Later, we provided the original service package to Digital China. The service was very good and the price was only half the price. There was no charge for overtime. In this way, Party A was slowly promoted. If you can't do it all, then I won't choose you. Anyway, I just spend money to solve the problem, but I am asked to distinguish the problems and spend money separately, so why don’t I choose an all-inclusive package? If I am so strong in operation and maintenance, do I still need to find you?
Therefore, current parties like to outsource projects to Alibaba Cloud and Huawei Cloud. These cloud vendors have industry solution architects, and all they make are solutions. The products below are all vague and can be used. . This is how involution comes out. But having said that, the market is not good or bad. The key is that you have to adapt to it, rather than judging whether the market is good or not based on your own technical experience. This kind of thinking is unacceptable.
Li Qipeng: Yes, if you work in different markets, you must respect the rules of different markets. This is how China uses this method to drive industrial development. It is also difficult to judge a market by whether it is good or bad. I went to inspect the Japanese market last month and found that the Japanese market was particularly difficult to enter. At that time, a company in my industry took a year to get an order worth 800,000 yuan from the Japanese market. Because it has POC, safety certification, etc. audits, the threshold is relatively high. However, once you eat this market, you can keep eating it, because they rarely take the initiative to change suppliers. Therefore, this is also a characteristic of the Japanese market. If we want to do something, we must respect its rules, both at home and abroad.
Brother Tou: That’s very well said. Qipeng also mentioned a point just now, that is, domestic integrators are relatively powerful, and many parties will treat you as an outsourcer. They don’t care whether your products are standard or not, and they will also mention a bunch of irrelevant requirements. Each project will be based on Money counts per head. What do you think about this phenomenon?
Li Qipeng: I personally feel that if you want to do a large-scale project in China, the requirements for the person in charge of the project are quite high. The person in charge may need to consider many aspects. One situation I have seen is that a project incubated a product and an industry was created through this product. There was a monitoring company like this before. It started with China Construction Bank, and then after improving its products, it was promoted to Shanghai Pudong Development Bank and other banks. Then I think this type is quite valuable, because the problem you solve is a general problem, and you abstract it into a product and achieve scale.
Of course, it’s more like what you said, Brother Tou, who doesn’t care about commercialization of your product at all, he just wants to solve his problem. Of course, there is no problem with this. After all, it is Party A who pays for it. In this case, the person in charge of the project may need to consider the investment issue, or how to invest in the project. One way is to directly focus on your own research and development and all in this project, and the other way is to find some partners to win this project together. Many companies may even find some temporary outsourcing first, because if you only want the head, then I will only give you the head, and then I will only make money from the head. But some companies may say I'm sorry, I won't make any money per head, and I won't do this project.
I think you still need to think clearly, because if you don't think clearly, it may have a greater impact on the company. It will impact your entire R&D system and even affect the normal iteration of the entire product. So when doing a project, you must think clearly about what kind of project you want to do, what kind of users you want to serve, and how you want to serve them. This user portrait must be clear, otherwise it will disrupt your rhythm.
Brother Tou: There is another question. As ordinary technical people like us, how do we monetize the technology we have?
Li Qipeng: I think joining a large company may be a better way. Because my own experience is like this. At that time, I was working in the Alibaba community and saw that many friends were still in a small company, but they were more active in our community and contributed a lot of code. This gave him experience and The experience was a great blessing, and later they all moved to big companies. From an ROI perspective, or from a revenue perspective, this is a very profitable thing. By contributing to well-known projects, you can improve your technology and influence, thereby improving your career. I think this is the most direct way. a way.
I have seen a lot of the second method, which is to do consulting after gaining technical experience, or to guide others to write code, publish books, courses, provide training, etc., and they have done quite well.
The last one is to start a business, but if a pure technical person wants to start a business, I suggest you join a start-up company first, if you can adapt. If the company is reliable, then when it develops you will have wealth and freedom. There is definitely a big difference between working in a start-up company and doing a detailed job in a large company. In a start-up company, if you want to build a Rolls-Royce, you may have to start with a bicycle and then build a bicycle. An electric car, then an Alto, then a BMW, and finally a Rolls-Royce. This is his process. I definitely don’t mean to give you a few years to build a Rolls-Royce. That would make it difficult for the company to survive in the market. But in this gradual development process, a technical person's technical vision and even business vision will be opened a lot. I have seen many people like this. They were purely technical people before. After joining this startup, they may be working on technology at first, then they may be product managers, and then they may be responsible for the entire pre-sales. His personal abilities will be greatly improved.
Of course, the choice also depends on the age group. For example, if I worked five years ago, I think you can still go to a big factory to have a look and experience it. But if you have been working for 5 to 10 years or even longer, I think joining a startup company is still a good choice. However, if you say that you want to start a commercial company as a partner or even a founder, then to be honest, I personally don’t particularly recommend it. Because if you are a partner, you will find that the first thing you focus on later is not technology, but users, financing, and the market. You have to spend a lot of energy to do these things. Some people may not Gotta love it. You may later find yourself becoming a salesperson while doing it. Of course, this does not mean that sales are not good, but he may not have liked it before, but he will be needed to do this job later. So the second thing to consider is whether this is something within your ability. Some people grow and change quickly and are very adaptable, so there is no problem.
Brother Tou: Last question, as an excellent open source project, if you want to try commercialization, what are the ways?
Li Qipeng: If a large manufacturer wants to achieve commercialization through open source, the most typical path now is to donate to a foundation and then gain customers by operating a community. This method is especially suitable for basic software. If it is a particularly vertical software, it may not be suitable. One way I personally agree more is to set a benchmark through open source and gain recognition from developers of certain benchmark companies. At this time, you may be able to copy it very quickly. For example, if a major manufacturer adopts my software, the second- and third-tier manufacturers below may also follow suit. Once your influence is established, it may be easier to seek payment.
But now, in 2023, at this point, it is more difficult for you to establish a project to start a business. First, the path to commercialization of open source is relatively long. You must first have a project, and then increase its influence through operating the community, and then release a commercial version to monetize it. This may mean that the startup company will take a year, two or even three years. There is no income every year, or you cannot support yourself. In the current employment environment, including the investment and financing environment, whether you can survive for two or three years is a very big question, and many people cannot survive it.
Also, speaking from my own experience, many users only use the open source version. They have never thought of using your commercial version, or they leave directly after you release the commercial version. This is very difficult. Difficult to convert. Some open source companies have succeeded in commercialization, not because they converted open source users into commercial users, but because after the project became influential, it affected users who did not use open source projects, thus achieving commercialization. This is an indirect impact that cannot be quantified. From a top-down perspective, it is difficult to formulate an assessment mechanism to let everyone know what is valuable and guide them to do what they do. So I think if it is SaaS, it will solve your problem of polishing the product better than open source, because there is data on it, and you can see how much they use it.
Finally, I think the greatest value of open source is standardization, such as Conflict, which is a standard for building big data platforms. Regardless of whether you use its open source version or buy its commercial version, you will think of it whenever you build a big data platform. Secondly, open source helps internationalization and gives Chinese companies an opportunity to go overseas or serve overseas customers, which is difficult for local closed-source software companies to do. I once thought that open source plus cloud was a good way to help Chinese software companies become companies that serve the world. Before there was no cloud, you had to build a local team if you wanted to serve overseas customers. Now you can just rely on AWS, which greatly saves costs.
Of course, open source has many benefits, and tomorrow will be wonderful. The challenge now is to see if you can survive until tomorrow. (laugh~)
The replay of this live broadcast is as follows, please scan the code to view it~
