How can we select the right users for user research?

The general idea is that the target user should be a typical user of the target product. This seems to be a very simple problem, but sometimes it is very complicated. Many Failed User studies are completely caused by improper user selection. It can be said that if the user selects a reasonable one, the entire study will be half successful.

The so-called "typical" is probably the people who use a product in the center of the normal distribution bell-shaped curve. Well, if we are now preparing to interview users who use their mobile phones to log on to the 139 mailbox and send emails, we find that most users who have used mobile phones to send emails only send three or four emails per month (assuming the situation is true ), if we find these people for interviews, they may provide very limited information about the mobile phone logon 139 email sending experience. On the contrary, if we look for users at the rightmost end of the bell-shaped curve who log on to their 139 email every day, the information we receive from them will be very rich and enlightening.

Therefore, the general principles for selecting users in user research (especially qualitative research) should be:Find users who have a rich experience (in the target product, product in the same domain, or product in the same domain.However, three cases need to be clarified here:

1) if the target product is not available and there cannot be users with experience, you should find users with experience in similar products;

2) if we aim to find out why users do not need a certain (class) product, we should find users who do not need to be killed;

3) For usability testing and eye movement research, it is often more valuable to select new users.

In addition, this principle hides a small organ: "Products in common fields ". What does it mean? Let's take a look at the classic practice of IDEO's user research: if they are doing preliminary user research for a top-level restaurant under planning, the users they are looking for Will not only attract customers who often patronize high-end restaurants, but may also go to the manager of the golf course or the manager of a luxury brand store, because these people know how to satisfy the needs of rich people from the bottom of their hearts. That is to say, the users we are looking for can not be limitedTarget ProductPeople with rich experience, and can be"Psychology/cultural genes" shared with target products"ProductSenior users or experts.

When looking for users, another common problem is: whether to subdivide user groups in advance, users are divided into several different user groups (usually divided by gender, age, occupation, education level, and other dimensions). Users are filtered based on these classifications. In fact, this is usually unnecessary. Here we need to clarify the difference between consumer segmentation in market research and user research in UCD. If market research is typically conducted based on the general attributes of consumers, the UCD field focuses more on the interaction process between users and products (including: users focus on the behavior and attitude of people in the interaction process, while the interaction design focuses on how the products behave and respond in the interaction process ), user-centered design can be called Action-centered design to some extent.

Of course, there is no objection to classifying users, but this classification must be completed after the user's research is completed, rather than before (otherwise the research has not been done yet, why do you know that your category is appropriate ?); It must be based on the behavior patterns and characteristics of the user's interaction with the target product, rather than just some attributes of the user. Here, I admit that factors such as gender, age, and occupation can significantly influence people's behaviors, but these factors are not decisive under specific product constraints, it is usually not a good classification basis. For example, the players of warcraft have both senior high school students and doctors and lawyers, but you may not be able to tell the difference from their behavior characteristics of playing Warcraft. Of course, I am not opposed to asking users about demographic parameters when recruiting users, but it is not for user classification, but to ensure a certain coverage and balance of samples.