An Introduction to Portrait in User Research Methodology

In the post titled “Summary of User Research Methodology” published last year, I have compared user research methods in two textbooks. Secondly, I will give an introduction to portrait in user research methodology. The following introduction is quoted from the article “User Portrait-Method” (abridged) or “用户画像-方法”.

Portrait is also called persona or personality. It is a virtual representative of a real user, a virtual user based on a deep understanding of real data. We use research to understand users, divide them into different types based on their goals, behaviors, and opinions, and then extract typical characteristics from each type, give them a name, a photo, some demographic elements, the description of the scene, etc. to form a portrait.

When we have multiple portraits, we need to consider the priority of portraits. In product design, meeting the needs of primary portraits should be firstly considered and then trying to meet the needs of secondary portraits without conflict. Of course, when a product is very complex, we may need to consider the priority of its portraits for different modules. For example, in a comprehensive shopping website, a female character is the primary portrait in the clothing section for women, but in the wear section for men, it becomes a secondary portrait.

The best practice is to conduct meticulous research and create a portrait of the product in the early stage of product development. However, in actual operation, many times you may think that a certain product can be done and just do it. There is a relatively large deviation between users and previously imagined users, and the product architecture designed based on the previously imagined users is difficult to carry the needs of actual users. At this time, the first task is still to define the target users of the product.

Portrait can be created through qualitative research in the early stage. Of course, if necessary, you can also verify the obtained portrait through quantitative research later. However, even if you want to create quantitative portrait, sufficient qualitative research in the early stage is also very important. In the interpretation of cluster analysis results or parameter adjustments, a full understanding of users can help us create meaningful portrait.

The creation of a portrait can be divided into the following steps:

Research Preparation and Data Collection

As with all research, we must first determine the type of interviewed users, design the research plan and the research outline.

The first question that arises is: Who are we looking for. Since the purpose of research is to create portrait, we should investigate as many different users as possible. Through brainstorming with colleagues in different departments to find out the various possible user types, we may get a list of conditions, or a user matrix as follows, and then we can invite users based on these conditions, three for each type. However, the choice of users should be more flexible. If we find that a certain type of user is missing during the research process, then add this type; or after we researched 20 companies and found that there is no new information, then we can cancel the remaining surveys.

When selecting research subjects, we should not forget other stakeholders besides the actual users of the product. For example, it is the wife who buys household fresheners, but the husband’s and children’s scent preferences will also affect the wife’s purchase decision; the business owner may not use or rarely use a certain product, but he is one of the key figures in the final purchase decision. So these people should be included in our research. For enterprise products, distributors are also our very important research objects. The research method to be adopted is mainly considered based on the research purpose, project time and funding.

The key to distinguish different user types is the user’s goal and motivation for using the product, past/present/future behavior, rather than demographic characteristics such as gender, age, and region. The survey outline is designed according to the actual situation of different products.

Affinity Diagram

Affinity diagram is a method for summarizing and sorting out a large amount of collected qualitative data such as facts, opinions or ideas according to their proximity.

Through the data collection in the previous stage, we have collected a lot of data. How to involve multiple people in the process of data analysis without missing the data? Affinity diagram is very suitable at this time. The advantage of this method is to allow a large number of the analysis process of qualitative information is visualized, which is convenient for everyone to work together and unified understanding. At the same time, affinity diagram produced can be conveniently used as a data basis for discussion in the next stage.

Portrait Framework

Through the affinity diagram, we have identified several types of enterprises, as well as the types of individual users in the enterprise. Next, we can describe the important characteristics of these companies and individuals to form a framework for portraits. In this step, we do not need to add descriptive details, just list the key content. Basic information can be described by scope. For example, the number of employees can be written as “less than 20”, and the specific number can be defined in the next portrait.

The purpose of this step is to quickly discuss with the rest of the team and collect feedback before the portrait of end users is output.


The next thing to do is to work with the product, market, and each group of leaders to complete the prioritization of portraits. When determining the priority of portraits, we can mainly consider from the following aspects:

(1) Frequency of use

(2) Market size

(3) Revenue potential

(4) Competitive advantage/strategy, etc.


The last step is to perfect the portrait. The main things we need to do are:

(1) Combine real data and select typical features to add to the portrait

(2) Add descriptive elements and scene descriptions to make the portrait full and true

(3) Concretize the scope and abstract descriptions in the portrait framework. For example, change the number of employees from “less than 20” to “15”

(4) Make portraits easy to remember, such as using names, iconic language, and a few simple key features, which can alleviate the memory burden of readers