Useless with research? --Reflections on the Practice of user research

It has been 5 years since the graduation internship to work on usability. In this record some of the author's view, and to discuss the share.

User research in the "user-centric" interface design method is very basic, but also a key link. But in practice, the author often heard the complaint is "user research input, time-consuming, and useless." There are certainly many reasons for this sound, I would like to discuss from the following four aspects.

Low quality of employees?

Some people say that the current availability of quality workers low, so they do not have the level of research, affecting the industry outside the view of the industry. I agree with this formulation, but I do not discuss it for the time being. At present, there are a number of employees of the survey report can be read, the qualifications of the employees of the background, such as a more detailed description. For the simplest example, in 2009, the company's top leaders sent out emails that said, "All designers have to write some html,css, even PHP." "Xi ' an Jiaotong University professor Li Le once said, do Man-machine interface design to write code." And the author will only be the simplest HTML, the basic requirements can not be achieved.

is the business value of user research difficult to gauge?

It is also said that the commercial value of user surveys is not easy to measure, so the value of research is not easily reflected. I do not agree with this. Will commercial value be higher without user surveys? This is a historical turning back. The past more than 50 years of computer hardware and software development has been a step in the way of learning the lessons of history. After some problems have been resolved, people tend not to feel any progress--just as people have become accustomed to using the mouse, how much change has happened to human-computer interaction compared to using a keyboard-how much time does it take if one step lets users spend 1 minutes a day with tens of millions of people using the software? How much bandwidth is wasted and how much server resources are wasted using Internet products? Go back to 50 and see what's going on. Welcome to talk about this issue. Secondly, it is difficult to estimate the value of its ideas and methods are not only user survey, for example, there is corporate culture, brand building. It is difficult to estimate precisely how much value corporate culture can create for a business. Even in the field of market research, the brand value assessment as an example, there are many models, and also continue to develop models, but this "difficult to estimate the value of" brand building is still constantly building projects, and is constantly being applied.

The impact of agile development on the use of research

The third problem is the field of software development, the author thinks that the idea of "agile development" brings some influence to the user research practice.

Figure 1: Agile Development (Source: http://wiki.mbalib.com/wiki/%E6%95%8F%E6%8D%B7%E5%BC%80%E5%8F%91)

In the summer of 08, Professor Li Le, an XI ' an Jiaotong university, said to Tencent Designer: "We have a user survey ahead of us to build a user model, and the user model is designed to help write design guidelines for software developers to refer to." Many of the people involved in software development are not familiar with the user, but he wants to cover all the user requirements according to the design guidelines you write. If you omit one, he may have omitted a structure, a function. In the end, the whole software is useless. ”

This situation can easily occur in a software development project that uses a "waterfall" development process. From the design to the development of product characteristics, to development, testing, interlocking, one step carelessly, the full plate is lost. The concept of "agile development" was raised in 2001. At present in Tencent, more and more product development began to adopt the "iterative" development process. It is understood that other Internet companies are also in use. The main approaches to this "iterative" development are: the whole process of product development is a series of iterative processes, fixed release time, release speed, some Internet products every half month issued a version, each provide some new product features, gray volume, a small number of users first try, and then collect this part of the user feedback, Quickly modify the version and then mass release the product.

The main impact of this "agile development" on user surveys and design is: Product release speed, but the investigation from the beginning of the project, the implementation of analysis, all need to spend a lot of time, often not catch up with the release of the rhythm; Through the gray volume, users can use the actual product feedback after the user information Instead of relying on a large number of previous analysis to predict the user's use of new features, the architecture is flexible, the cost of modifying the code is low, allowing development errors, do not rely on the preliminary survey results as a guide.

In essence, agile development is not antithetical to user surveys. "The core of Agile thinking is human communication," says ThoughtWorks, a consulting firm that aggressively promotes agile development processes in China. The question of demand is actually a matter of communication. Business analysts want to communicate with customers to find out what the customer really needs, why do they need these things? Business value is the ultimate goal of business analyst attention, with the goal of pointing can not lose direction. The ultimate goal of communicating with customers is to tap into the customer's business objectives. Maybe people will often have this experience-the customer said I want this function, I want to how how kind, at this time to pay special attention to he said that these things may not be the real demand, business analyst needs to ask the customer why, dig out his real goal. "One of the typical agile process models is called Extreme Programming (Extreme programming), which manages requirements through story cards (Story card). This kind of story card can be said to be almost a scene analysis, task analysis of another performance. "Agile development" is not no longer focus on users, but from the constant trial and error to obtain user feedback, thereby improving the product, in fact, is a "user-centric" another performance.

Figure 2: Story Card (Source: http://www.scissor.com/resources/teamroom/)

In this sense, "agile development" itself is "people-oriented" for programmers, because it allows programmers to make mistakes, emphasize the communication between developers, and even encourage programmers to communicate with users, hoping to make mechanical programming into enjoyable activities.

Therefore, the interface designer's attitude towards "agile development" should not be inconsistent and should seek solutions. The author's team is trying to solve the problem with tools and processes

We shorten the survey time by means of tools. In user interviews, it takes time to invite users, make records, and write analysis reports. In this respect, the author team is to establish a voluntary access to the user database, in the future when necessary to extract users; During the research process, invite project team members including development, designers, product managers to visit and observe users, the first time to share useful information, reduce document writing, only to write the necessary documents. Questionnaires, the production of the problem is very time-consuming, the company has developed a special questionnaire production and delivery system, from the production to the user to the fastest within a few hours to complete (not including design questionnaires). Data analysis phase, using SPSS analysis, the common analysis code is sorted out, the next statistical analysis can continue to use, reduce the analysis time. Shorten the execution time of the survey itself by using tools. Recently, a team of investigators received the demand to investigate the blind's opinion of the product, from receiving the demand to finding the blind to visiting the report, for a week, and for other projects at the same time, it took only 3 working days to complete the interview project.

We solve the conflict between long time surveys and agile processes through processes: the results of a user survey can be written in a number of "story cards", each development, only one part of the "story card", through a few iterations of the development cycle to complete a survey of the "story card" development, during which the new user survey can continue.

In addition, it is ultimately necessary to analyze the backend data generated by the user's use, which can still be the work of the investigator responsible for designing the survey.

Good "reference", light research

The fourth question, the comparison essence, is because the domestic enterprise likes "imitates", or says "the reference".

"Draw on" the question, will have the very big influence to the user survey. Some teams do not want to invest in research and design, the product requirements and design guidelines according to other people's molding products to write on the line. The team that wants to do research, has the innovation, designs the product, also has the very big risk by other team "draws" to form own product. (China's patent status on the interface design does not have substantial protection, specific performance: first-person interface in China often can only apply for appearance patents, is a static interface, not really suitable for the protection of Man-machine interface design; second, the appearance of patents are difficult to obtain, so sometimes the application is a technology patent, Other groups will not infringe a patent if it is implemented in a technical scheme, and once again, the lawsuit will be delayed for 2-3 years. ）

"Reference" is often the other cultural background, such as the successful operation of foreign products. However, culture, national conditions, different stages, direct "reference" may not be successful, even in Tencent, there are such cases. Product success, but also to do the target user survey and analysis.

"Draw on" is definitely not a real shortcut, it is not necessarily to achieve commercial success.

Therefore, on the fourth question, the first "reference" still exists now, or can not be separated from the user survey, to different cultures, national conditions, stages to make analysis and judgment to help product success, and secondly "reference" is not a shortcut, will not be long-term existence. When "borrowed" cannot be a reliable method, it will surely return to the path of innovation from the survey analysis.

Finally, an extra question. The relationship between designer and product manager has been much discussed in the Internet industry. The author's view is that the Man-machine interface design, in fact, and product manager's work overlap, so learn the Man-machine interface design after two positions can apply. Already in the design post, there is no problem, you may consider yourself as a special skill product manager. Interface designers have just come to the post, often with product managers conflict, that the reason for the opposition is that they are from the user point of view, product Manager is not. It is likely that the product manager will know more information and consider more things. To let the interface designer stand on the product manager's point of view, the first-time designers often think that from the user point of view is their full responsibility, adhere to from a point of view, the lack of the overall situation. You may as well start by seeing yourself as a product manager with special skills.
(This article originates from Tencent CDC Blog, please indicate the source when reprint)