Opening bricks: how to develop Internet projects

Document directory

• Internet project development is different from traditional software project development
• How to conduct data analysis and user behavior research
• Okay. Let's take a specific example.
• Recently recommended articles

Unlike other software projects, the development of Internet projects has its unique characteristics.

Internet project development is different from traditional software project development

Different from software development companies that require customization. The customer's requirements are clear. A group of people can ask, what kind of software do you want us to make? Internet projects are faced with countless uncertain people, and the needs of each person are unclear and different. How to extract the most important requirements and continuously revise the requirements and develop them is an important issue.

For Internet projects, you cannot write the requirements and functional modules in 1, 2, and 3 at the beginning and put them in the project documentation. Because the requirements you have written are not necessarily correct. At least you cannot say that these requirements are correct now or when the project is launched.

Therefore, using waterfall-style project development management like traditional software development is basically a dead end. We must use a more agile project development method. I personally prefer the function module iteration method. The most basic function module is provided first, and then the following project requirements and function modules are adjusted in a timely manner. It must be remembered that policy makers should only observe and analyze the market at the beginning to provide the initial needs. The subsequent requirements are not determined by the head of the project at the beginning, instead, we observe the data of projects that have been launched and constantly determine the user behavior after analysis.

The most important feature of Internet projects is: fast speed, short time, and Website user guidance requirements.

How to conduct data analysis and user behavior research

As mentioned above, the subsequent requirements and development should be based on the access data and behavior analysis of current users. Therefore, data analysis and user behavior analysis are very important.

First, you need to accumulate your user samples. The more different and real users you have, the more accurate data analysis and user behavior analysis are. Therefore, the initial task is to expand the visitor of your project to expand the user sample of your project, because only in this way can your analysis be scientific and accurate.

This involves the operation of the project. In this case, you can access an Internet project I recently worked on: xinhujiang English website, which may be inspired.

Yes, I think there will be a lot of people who have clicked the link just now. In fact, this is actually a small means of website operation, and it is not worth mentioning. However, you need to know that website traffic is often accumulated in this way. Don't count on the huge traffic that the website traffic can bring through a certain means. Even if there is such a means, the user quality brought about by such means will not be very high. The improvement of website traffic should be in the form of "wanchuan huihai" in multiple ways.

It doesn't matter if you don't know. After that, I will explain how to use data for user behavior analysis based on actual operation data, you will be interested in this "very concise English learning website with excellent reading experience.

When talking about data analysis and user behavior analysis, the first thing we need is data: access data, statistical data, and so on. I think everyone is a programmer and there is nothing to say about how to get the data. Site statistics. You can use programs to record and analyze logs. The new Hujiang English website uses a combination of Site statistics and program statistics. We will record what users click on the website to analyze what they like and what they don't like. We record page jumps to analyze user access paths and user habits.

Okay. Let's take a specific example.

Maybe the above things are too doggy and boring. Well, let's look at a real and interesting example.

To protect company data confidentiality, the absolute value of the data is corrected to a certain extent while the comparison value is true.

Top 10 users prefer to click

Based on the analysis of background data, we know that the top 10 articles in the sidebar of the inner page are the ones that users prefer to click. However, almost all users' clicks are concentrated in the first tab. Almost all users only click the ranking article in the first tab, rather than clicking the other two. At the beginning, our three tabs were "weekly ranking", "Monthly ranking", and "Total ranking". We thought it was possible that the three tabs were all sorted by time, you are not interested in the other two time rankings. Therefore, we changed to "ranking by number of reading", "ranking by number of recommendations", and "ranking by number of comments ". However, the effect is still not good. The second and third tabs are still empty. However, when we change "click to switch tab" to "move the mouse to change tab", the effect is completely different. The number of clicks on the first tab has not decreased, and the second and third tabs have been greatly improved. The time comparison diagram is as follows:

Article reading ranking list click to compare




Article recommendation rankings




Rankings of comments pushed by Articles

From the data analysis, we can understand that "sliding the mouse to switch the tab" is more in line with the user's habits than "clicking and switching the tab. Of course, this is just a small example. However, the importance of Internet projects for data analysis and user behavior can be well described.

In fact, the behavior of Internet users can also be guided. Of course, this is another topic.

This article is actually a summary of work, but it says, it feels good to share it with you. Because it is a work summary, many contexts are omitted. This may cause some misunderstandings. Please forgive me. Haha ~~

Recently recommended articles

Gchart: full introduction to jquery components based on google chart APIs: 1. Getting Started

IxEdit silly JavaScript development tools (download, Chinese version, video tutorial)