leaf

The company had to disband our entire software development department due to a transformation, and as the head of the development department, I also had to resign. As I leave, I want to summarize my feelings about working at a startup for a year.

1. Interview
   At the end of January 2014, after the company held its annual meeting, I resigned. I organized my resume and prepared to start looking for a new job. By chance, I encountered the current technical head of the company recruiting people in a QQ group. After a brief chat, he invited me to meet him in person to discuss further. The meeting place was a café. After asking a few technical questions, he felt that I met his requirements and then started discussing salary. He offered a pretty good salary of 15K after tax, which tempted me. I also asked about the company situation, and he said it was newly established, mainly focusing on online education products. The product requirement documents, development environment, and project development architecture were all prepared in advance, so we could start development immediately. That was the interview process. I didn't agree on the spot and said I would give him a response after a week. A week later, he called to ask how my consideration was going. At that time, I had another offer from a different company, but the salary wasn't as high as his, so I chose the higher-paying company and agreed to join after the New Year.

2. Joining the Company
   On the morning of March 2, 2015, I officially reported to the company. Upon entering, I was stunned to find that the company didn't even have computers or internet set up. My first reaction was that I felt deceived. I asked him why the company had nothing, and he explained that the office location was just rented, and the desks were set up only a couple of days ago, so they hadn't had time to buy computers or set up the internet yet, but everything would be ready in a few days. I believed him. Gradually, other new hires also reported to the company, and soon the boss arrived and organized a meeting to introduce the company's situation. At that point, I learned that there were four partners in the company: two investors, a man and a woman, a female product manager, and him as the technical partner. The two investors didn't understand technology and weren't from our industry; the male investor was a lawyer, and the female investor was in finance. I later learned that the boss had spent a lot of money to poach an entire development team from another company, which consisted of seven people who had previously worked on a food ordering app. The team included members specializing in JavaWeb, iOS, Android, and database design, making it a relatively complete technical team. Three members of the team officially started working that day, while the other three had to wait two weeks to leave their previous jobs. The team leader had reportedly been developing for ten years and was considered an architect-level figure in my view. On my first day, due to the lack of computers and internet, I couldn't do anything and could only listen to him talk about the product requirements. This was quite frustrating; there wasn't even a document for the product requirements. He only sketched out the things and functional modules on a whiteboard. On my first day, I felt everything was unreliable. I privately asked someone from the technical team, and they also found it strange, saying they had never seen such a software company. Everyone said they would wait and see how things went; if it didn't work out, they would leave immediately.

3. Buying Computers
   After the third day in the company, the internet was set up, and the first batch of computers was purchased. Upon checking the computer specifications, I found they were completely unsuitable for development. They had bought Lenovo's home version, with only 4GB of RAM and no upgrade options, and the CPU was inadequate, making it very laggy. He was responsible for the computer selection, and later I suggested whether we could return this batch and choose other brands. These computers were really not suitable for development. He then let me select a batch of suitable development computers, priced around 3000~4000. So, my first task upon entering the company was not to do development but to help buy computers. I chose Dell desktop computers on JD.com and bought a few. The previous batch of computers was left for others, and all developers used Dell for development.

4. A Small Incident
   About a week after joining the company, for some unknown reason, the team leader of the technical team, who had been poached at great expense, had a conflict with the boss. The boss called him to the company and harshly criticized him in front of everyone in the meeting room. They fell out on the spot, reportedly because the team leader felt the technical head was unreliable and didn't understand anything, suggesting the boss replace him. The boss believed he was reliable and very capable, trusting him and unwilling to make a change. After the fallout, the technical leader did not come to work as previously stated after two weeks, and the remaining three team members also did not show up. Thus, the technical development team consisted of the three who had come earlier, plus me and another person, totaling five. Among the five, two of us were working on JavaWeb backend, one was designing the database, and another was doing Android development. He felt there were too few Android developers, so he asked me to switch to Android development. I had never done Android development before, but I was quite interested in it, so I agreed to switch.

5. Product Development Process
   In the second week after joining the company, we hired another Android developer, expanding the technical team to six people: three working on Android development and three on JavaWeb backend development. Without any product requirement documents, he asked us to start development on an online English education product based on Android tablets. The product consisted of several apps: 【趣学】, 【英语应用】, 【阅读应用】, 【益智游戏】, 【个人中心】, 【应用商店】, 【视频应用】, and 【Laucher 桌面】. Perhaps because there was no initial planning, we ended up creating a large and comprehensive product, which caused subsequent development delays, failing to meet the expected three-month timeline for product launch and market entry. I will elaborate on this issue later. He assigned me to develop the application store, but the task assignment was only verbal, with no use of professional project management software like ZenTao or Project to arrange development tasks. There was no requirement document for developing this application store, no UI design, and no project development framework; everything started from scratch. He told me to model it after Apple's application store. The product manager had an iPad, so I borrowed it daily to look at it and replicate it.

From setting up the Android development environment to learning while developing, there were many bumps along the way, and the pressure was quite high. By mid-April, I had created a rough version and showed it to him. He was quite satisfied. At that point, he told me that since the application store was nearly done, the Java backend team was short-staffed, and I should help with that. So, I helped develop the JavaWeb backend, mainly working on the app's data interface development. Our app's backend was primarily developed using Struts2, Hibernate, and Spring, which I was familiar with, so it was relatively easy for me.

By mid-May, the first version of my application store was completed, but disaster struck. Due to the lack of requirement documents earlier, the product manager had not specified what it should look like. Now, the product manager designed it and asked me to revise it according to her design. Fine, I made significant changes according to her requirements. After finishing, I showed it to the product manager, and just after the UI was modified, she told me the requirements had changed again and it couldn't be like this anymore; it needed another revision. Other colleagues also said their work needed to be revised according to the new requirements, which led to a loss of control over the development progress. Everyone had been working hard for months, and the originally planned product launch in June was delayed to July due to constant changes in requirements.

By mid-July, most of the related apps were nearly done, and we demonstrated them to the boss. After seeing them, the boss thought they were acceptable and planned to launch them to the market for validation. In August, a colleague and I went to Shenzhen to deliver our completed apps to a tablet manufacturer, who helped us install the apps on the tablets. We purchased 100 tablets from the manufacturer. After bringing the tablets back, we conducted over a week of functional and business testing, only to discover numerous issues. I wasn't sure if it was due to hardware quality problems or if the ROM had been rooted, but our apps were unstable after installation, frequently freezing and crashing, leading to a poor user experience. The boss played with them a few times and was very dissatisfied, feeling that such products were completely unusable for market launch. He asked us to resolve these issues before going to market, which delayed the launch time until September. It took a full six months from March to September to finalize the product, and the boss was very disappointed, growing increasingly dissatisfied with both him and the product manager, and began to question his capabilities.

6. Starting Over
   After the product was completed, we began to enter the market. Our marketing strategy was to sell tablets installed with our apps to users, making a small profit on the tablets, and then earning money from the services provided by the apps. The ideal was beautiful, but the reality was harsh. We didn't sell a single tablet, and the learning content within the tablets didn't attract users, so no one was willing to buy. Thus, the product we had worked on for half a year became worthless. Later, for some unknown reason, there was a significant personnel change in the company. The boss let some people he deemed unreliable leave voluntarily, including him and the product manager. The boss then spoke to me, hoping I could take his place, rebuild the development team, and completely overturn the previous developments, starting over according to his vision.

Starting in October, I officially took over the entire development department. I reassembled the team, keeping some who were willing to continue and work overtime unconditionally, while those unwilling to continue were let go. The development team was left with five people. After another half month, three colleagues left the development team, leaving only three of us: one designer, one Android developer, and I was responsible for the JavaWeb side.

We summarized the mistakes made previously and unanimously agreed that the earlier product was too broad and lacked focus, trying to do everything without emphasizing key points, resulting in each app lacking distinctiveness and not being well-executed. Among our products, only the 【趣学】 app was of significant value, while the others were merely supplementary. Realizing that the 【趣学】 app was our main focus, we dedicated all our time and energy to developing it. After multiple discussions with the boss, we finalized the specific requirements. I wrote a detailed requirement document, established a development plan and timeline, and we restarted development. The designer redesigned according to the requirement document, and the Android side brought over some previously completed components, while I completely discarded the entire backend development framework and rebuilt it according to the new requirements. It took about ten days, and our new version of 【趣学】 was completed. I demonstrated it to the boss, who was quite satisfied, saying this was what she wanted.

7. Operations and Promotion
   After the new version of 【趣学】 was completed, the boss began recruiting professional English teachers and foreign teachers to record English learning videos, which were then uniformly managed through the backend management system and displayed in the 【趣学】 app. Next came the app promotion; the boss hired several people to form an operations team responsible for the app's operation and promotion. During this period, we made modifications and improvements based on feedback from the operations team.

Around mid-November, the boss saw a new concept English app and thought it was quite good, but that app lacked a learning statistics feature. So, she asked us to create a similar one, adding a backend learning statistics feature. The development cycle for the new concept English app took about 20 days, and we created a nearly identical app with the added backend statistics functionality to track users' learning progress.

Once the new concept English app was completed, we began considering how to monetize the app. Our profit model primarily relied on the services provided by the app. We negotiated partnerships with several English training institutions, where they would handle student recruitment while we provided tablets and apps. Students recruited by the training institutions would use our tablets to learn English, and our app's backend management system conducted detailed data statistics and analysis of their learning progress. In addition to collaborating with training institutions, we also set up our own training classes and recruited students independently, hiring part-time teachers to conduct classes. During this period, we reached cooperation agreements with several training institutions and established a few training classes. However, due to issues with student recruitment, the number of students was relatively small. To attract more students, we distributed flyers, held free trial classes, and promoted through WeChat, employing various promotional methods, but the results were mediocre.

Time flew by, and 2015 passed just like that. The company held its annual meeting on January 30, 2015, distributing project bonuses and having a dinner. In the annual meeting summary, we reviewed the mistakes made in 2015 and made specific plans for 2016, aiming to sign contracts with 50 training institutions and develop a teacher-side app and a parent-side WeChat app. After the annual meeting, the company went on holiday.

In 2015, the company had no profits at all and burned through over 3 million.

8. Team Disbandment
   At the beginning of 2016, the company officially resumed work on the 8th. A new year brought new hopes, and everyone was full of confidence, hoping to smoothly promote the product as planned before the New Year and make the company profitable. However, reality was still harsh. Our promotion efforts and partnerships with training institutions did not progress well. From February 15 to March 15, the company invested considerable manpower and effort into promotion, opening tutoring classes and inviting foreign teachers for free trial classes, but the results were still unsatisfactory. The costs incurred by the company far exceeded the income. On March 15, without any warning, the boss called the entire operations team into the meeting room and announced that the company no longer intended to pursue the education business, disbanding the entire operations team. That afternoon, they settled their salaries and let them go. At 5 PM, the boss called me into the meeting room for a discussion, explaining the company's current situation and her thoughts. She felt that making money in the education sector was too difficult and did not want to invest more time and money into it. The situation the company was in was not the fault of the software development department; the apps we developed were completely fine. The main responsibility lay with her and the operations team, as we had not clearly defined a profit model. The company temporarily did not need the software development department and currently had no software-related tasks for us. We were told to complete the remaining work and then look for new jobs. Thus, our entire software development department was disbanded.

9. Insights
   Over the past year at the startup, my deep realization is that entrepreneurship must be in a familiar industry, with a clear idea of what to do and finding reliable people who can help us achieve it. I summarized the reasons for our company's failure:

10. Both bosses were not from the IT industry and had limited understanding of the internet sector, making it difficult to grasp market direction.

11. The other two partners found through friend recommendations were also not particularly reliable, lacking in capability and experience, and similarly unable to grasp the entire product.

12. The product we aimed to create was merely the idea of the two bosses, without any market demand research.

13. The company's management was quite chaotic, with unclear job responsibilities and insufficient execution from the bosses. Many things were only discussed verbally without concrete implementation, and some decisions were never finalized.

14. The early development process was severely unregulated, lacking a specific development plan, adopting a parallel execution of product design and development, and inadequate communication between the product and development departments, resulting in a significant gap between the developed product and the initial vision, leading to constant revisions and repeated delays in the development cycle, missing the optimal launch time for the product.

15. We did not clearly define the product's core competitiveness and highlights. Initially, we aimed to create a large and comprehensive product without focusing on key points, wasting a lot of time and energy on irrelevant apps.

16. Some preliminary preparations were also inadequate, with delays in tablet selection and ROM customization, which were only hurriedly finalized right before the product launch.

17. Gains
    Over the past year at the company, I have gained quite a bit. In terms of development skills, I mastered Android development, improved my WeChat development, and enhanced my JavaWeb development skills. I also participated in some decision-making regarding the company's operations and management, expressed my opinions and views, and transitioned from an ordinary developer to a project manager and department head. As a product development leader, you must have the ability to assess the project's development cycle and identify its challenges, allowing you to formulate a development plan based on the project's situation. You must also understand the personal capabilities of team members well and assign tasks to the right people for what they do best.

18. Year-End Achievements

19. App Backend Management System
    The app's backend management system is my masterpiece, developed using Jfinal2.0+EasyUI. All of the company's apps currently rely on this backend for operation.

20. App Results