How to make valuable user research? Talk about the method

1. Why do user research?

User research is systematically understood through scientific methodsUser behavior and attitudesunderstandUser experience, pain points, and needsThe process ultimately provides a subjective and objective basis for product decision-making.

User research is the key bridge between products and users

Any successful product requires a deep understanding of the user, which helps the team to:

1. See the real needs of users clearly (don’t be fooled by superficial needs)
2. Validate product hypotheses (avoid developing features that no one else uses)
3. Continuously optimize the experience (find the pain points that make users scold)

When a team experiences the following symptoms, it means that user research is needed:

1. “I think users should…”, the team should stop speculating and ask real users
2. “Competing products have this function…”, the team should first figure out whether users need it
3. “How cool this design is…”, but cool does not mean easy to use
4. “The data looks good but…”, but the people behind the numbers are more important

The focus of research at different stages is different

Product concept period: find the direction, research what is the biggest headache for users?

1. Core goal: dig out pain points and find opportunities
2. Key value: Distinguish between surface needs (faster horses) and intrinsic needs (more efficient means of transportation)

Product design period: anti-stepping pit, will users use this design?

1. Core goal: understand the scenario and optimize the plan
2. Key value: Ensure that the design solution is in line with the user’s thinking habits and avoid making products that look beautiful but are difficult to use to cry

Product development period: risk avoidance, is the direction of research and development right?

1. Core objectives: test hypotheses and control risks
2. Key value: Continuously verify whether the direction is correct and correct the route in a timely manner (e.g., find that the actual demand of community apps is content consumption rather than social networking)

Product operation period: continuous change, how to do better and better research?

1. Core goal: upgrade experience, fast iteration
2. Key value: Through user feedback, continuous iteration to polish good products into explosive products

2. What exactly is user research doing?

Research who the users are

Basic information：

1. Age, gender, residency, how much money is earned, what education
2. Is it an office worker or a student, a few people in the family, what kind of circle is it?
3. Whether your brain turns fast or slow, what you like and hate, and what your personality

Professionalism：

1. Is it a novice or an old driver
2. Willing to try new to the old man or conservative
3. How long has a product been used and whether it will play advanced features

interpersonal relationship：

1. Whose opinion to listen to (influencer/friend/family)
2. Whoever pays for it (for example, parents buy for their children)
3. Whether the company culture has had an impact on him

Study user behavior

Use paths：

1. The ideal steps to use vs. how to mess up the actual situation
2. Which step is easiest to give up (e.g. the shopping cart suddenly stops buying)
3. How to match your favorite (such as searching first and then comparing prices)

Usage habits/frequency：

1. Use it daily or occasionally
2. What functions are mainly used and which functions are used to eat ashes
3. What is the difference between weekdays and weekends?

Abnormal operation：

1. How to bypass difficult features (such as taking screenshots instead of sharing)
2. Develop unexpected uses (e.g., using a shopping cart as a favorite)
3. He said he didn’t want to be very honest with his body (for example, it was too expensive but he used it every day)

Study user attitudes

Cognitive level：

1. What is the use of the product?
2. What do you value most when buying things (cheap or quality)
3. How much is the difference between the expected and actual experience

Emotional level：

1. When used, it is scolding or praise
2. The more you use it, the more you like it, or the more you use it, the more annoying it becomes
3. Which moment made him want to recommend it to others

Real needs：

1. Say what you want (want to save more electricity)
2. What do you actually want (it turns out to be slow to charge)
3. The ultimate goal (available anytime, anywhere)

Study the user environment

Physical environment：

1. The product is not easy to use
2. Differences in the use of different terminals
3. What is the difference between usage in private spaces (home) and public spaces (company/road).

social environment：

1. How will they perform in the presence of others
2. Are there two types of personalities for going to work and for getting off work?
3. Differences in usage due to regional cultural differences (e.g., south and north)

Digital environment：

1. The mobile phone and computer are not stuck when switching
2. Is the traffic sufficient and the network speed is not fast?
3. How to restrict the use of app store rules

The contradiction between user behavior and attitude often hides real needs

Cognitive dissonance conflict (self-comforting users)

1. Performance: Users change their attitude to rationalize their behavior (they automatically think it is good after buying something)
2. Case: After purchasing luxury goods, users’ brand evaluation is automatically improved, even if they are initially dissatisfied (after buying an expensive bag, it is worth it even if there are obvious defects)

Social approval contradiction (say one thing and do another)

1. Performance: Public attitude deviates from private behavior, claims to be different from actual choices, users give high ratings but use them infrequently (they sound good, but they don’t do it)
2. Case: Moments post fitness, in fact, it moves twice a month; 70% of users claim to value environmental protection, but only 30% are willing to pay a premium for environmentally friendly products

Group polarization contradiction (with the flow type of users)

1. Performance: Individual attitude is mild, extreme after group discussion (easy to be biased during discussion)
2. Case: In the focus group, I thought it was okay, but I listened to others say it was not good, and I followed the opposition, and the neutral users finally supported the radical plan (group polarization effect)

Habitual inert contradiction (lazy modifier user)

1. Performance: Pursuing innovation in attitude, relying on old models in behavior (knowing the new good but still using the old)
2. Case: The user supports the new interface design, but the old mode is still operated after the update

Time decay contradiction (three-minute heat type user)

1. Performance: short-term attitude is inconsistent with long-term behavior (start to say yes and then don’t use it)
2. Case: The new feature is great when it first came out, but no one uses it after three months

Cost-sensitive contradictions (stingy users)

1. Performance: Attitude support is separated from paid behavior (say support but don’t want to spend money)
2. Case: 80% of users support the “de-ad” function, but only 2% purchase members

Dependency-resistance contradiction (love-hate users)

1. Performance: Enjoy the pleasure of instant gratification, but also have the anxiety of losing autonomy (use while scolding)
2. Case: Complaining about Douyin every day is a waste of time, but brushing for 3 hours a day, but the negative evaluation user is a high-frequency user (there may be an addictive design)

Scene fragmentation contradiction (laboratory good baby type user)

1. Performance: Difference between laboratory attitude and real scene behavior (good performance during testing, actual use of stuffing)
2. Case: The task was completed 100% in the usability test of the experimental environment, but only 40% was successful in actual use

Core principles: The record must be complete, and the analysis must be thorough

Retention of Complete Evidence (Records)

1. Completely preserve audio and video materials, and organize key content into transcripts
2. Benefits: (1) Those who did not participate can quickly understand; (2) Avoid “selective memory” and fall into the pit of “self-evidence”

Exclude subjective bias (analysis)

1. Don’t just stare at the evidence that supports your idea
2. Don’t blindly trust your memory, memory may be distorted, but records will not
3. Don’t expect users to give answers directly
4. To see the essence through phenomena: behavioral data + attitude expression + environmental factors = real needs

Teamwork (participation)

1. Avoid “one person research, all blind followers”
2. It is recommended that team members take turns participating in the research, reviewing the original materials together, and finally collectively discussing and reaching conclusions

Step1: Clarify the purpose and problem

User Research Step2: Select the research object

User research needsScientific stratification

User research must ensure that the sample can truly reflect the user ecology of the product, and there are three core requirements:

1. Representativeness: Covers key user types
2. Randomness: Avoid subjective selection bias
3. Pertinence: Adjust the dimension weight according to product characteristics

Hierarchy from the user attribute dimension

Dimension 1: Basic Demographic Characteristics (Hard Indicators) Demographic characteristics are the basic dimension of user research, including objective indicators such as gender, age, income, and education.

1. Different products need to focus on different key characteristics: high-end appliances focus on income level, and fashion products focus on gender and age.
2. Be wary of the misunderstanding of “high-end users”. Users do not have a high and low end, high-income and highly educated people do not necessarily have to buy high-end products, luxury home appliances are actually operated by nannies, and complications become cumbersome.
3. The age segment should be in line with the actual scenario, and the “25-34 year old newlywed home appliance consumer group” is more meaningful than the “21-30 year old pan-age group”.

Dimension 2: Behavioral motivation system (why use) User behavior motivation presents a three-level structure: deep motivation → specific goals → operational tasks. This layering method can effectively distinguish between user types, such as dividing refrigerator buyers into practical function-oriented, the former focusing on freezing capacity, and the latter pursuing brand grade. Dimension 3: Cultural lifestyle (how to live) The cultural lifestyle dimension reflects significant regional and intergenerational differences. East Asia prefers beauty features, while European users prefer natural or nostalgic styles. There are also obvious subcultural differences within China, and the consumption concepts and lifestyles of different generational groups are very different, with the post-70s being pragmatic, the post-90s focusing on individuality, and the post-00s having an emerging consumption view. Culture is a pluralistic and complex concept, which will take on various forms with regional differences and time evolution, mainly including the following levels:

1. Living customs: behavioral patterns and traditional habits in daily life
2. Values: Belief systems and moral standards shared by the group
3. Aesthetic orientation: the perception and expression of beauty
4. Taboo norms: socially agreed behavioral boundaries

Dimension 4: Geographical and environmental factors (where to use) Geographical characteristics determine the regional differences in product function design, which is an important basis for product localization and improvement. It is mainly reflected in the following aspects:

1. Differences in climatic conditions: Jiangsu, Zhejiang and Shanghai rainy seasons are humid, winters are cold, the demand for dehumidifiers is high, and electric heaters are needed in winter, but dry is rejected; The northwest region is dry and rainless, with a large temperature difference between day and night, so humidifiers are essential, and air conditioners need to take into account the refrigeration/heating efficiency.
2. Differences in eating habits: Sichuan and Chongqing regions love spicy, many hot pot bases, refrigerators need to store hot pot ingredients in a large-capacity refrigerator, and range hoods need to strongly exhaust spicy flavors; Guangdong has a soup culture, fresh ingredients, fresh seafood storage area in the refrigerator, and reservation function for electric stew pots.
3. Economic characteristics: Hainan/Xishuangbanna is a tropical fruit producing area, and the refrigerator needs to be designed with a tropical fruit preservation mode (anti-frostbite) and a high-temperature resistance design; The black soil in Northeast China is rich in grain crops (soybeans/corn), and it is necessary to have a moisture-proof and mildew-proof storage scheme.

Stratify from the dimension of product service objects

C-end user stratification system: C-end products need to focus on user attributes and behavioral value dimensions

1. Underlying attribute dimensions：
   1. Demographic characteristics: including but not limited to age (18-24 years old Z generation / 35-44 years old Mesozoic generation), gender, region (first-tier cities/sinking markets), income level and education level and other hard indicators
   2. Social attribute characteristics: Covers soft dimensions such as family roles (new mothers/empty nesters), professional characteristics (996 office workers/digital nomads), and lifestyle (fitness experts/home culture groups).
2. Behavioral value dimension：
   1. Active users (3 active ≥ per week): Help validate the product’s core value proposition and benefits
   2. Sleeping users (30 days of inactivity): Focus on analyzing churn drivers and recall opportunities
   3. High-value users (ARPU/average revenue per user >industry 80th): guide the design of value-added services and membership systems
   4. Potential value users (similar characteristics among new users): Optimize your acquisition strategy and conversion funnel

【B-end user stratification system】B-end products need to focus on the differences in decision-making chain roles and functional requirements

1. Decision-making hierarchy：
   1. Executive layer (front-line business personnel): Focus on pain points and efficiency bottlenecks in daily operation processes
   2. Management (department head): Pay attention to system integration needs and team collaboration pain points
   3. Decision-making level (business owner): Focus on return on investment (ROI) and strategic alignment
2. Functional dimension coverage：
   1. Key functional roles such as business units (sales/customer service) and support (IT/finance) should be covered to ensure that the collected requirements can reflect real business scenarios.

Research Step3: Select the research method

Use research methods to classify dimensions

User research methods can be divided into four categories by two dimensions, the first dimension is qualitative and quantitative, and the second dimension is to study user attitude and user behavior.

1. In the first dimension, qualitative research is used to answer “WhyTo solve this problem? “And”HowQuantitative research is used to collect specific data such as “proportion, frequency, distribution, and degree”.
2. In the second dimension, user attitude refers to “What the user said“, User Behavior refers to “What the user did”。 In user attitude research, participants often hide their true thoughts, resulting in data distortion, and they need to avoid blindly accepting user expressions and directly using them for decision-making, which is a typical “self-deceptive research”.

In addition to this, there is an additional third dimension to consider”The context in which users use the product”。 User is (1) in authenticliving/working environment where there is no human intervention from the investigator, or (2) the user as established by the investigatorscriptExperience using the product, or (3) the userNot usedor completelyI don’t knowThe product.

The research method is specifically introduced

User Research Step4: Visualize the results of the user research

Presentation method 1: Storyboard

Storyboarding is a visual storytelling method that transforms user research data, product usage scenarios, or design concepts into continuous images or images to simulate the interaction process between real users and products/services. It is similar to a movie storyboard, using a combination of graphics and text to present user behavior, pain points, emotional changes, and solutions. How to make an effective storyboard?

1. Choose from user interviews/diary studiesHigh-frequency or high-pain point scenarios, showing the environment in which the user uses the product, including details such as time, place, equipment, and surrounding interference
2. Describe user operation steps through storyboarding, embed user quotes, behavior data, and annotate key card points
3. Visualization Considerations: (1) Hand drawn sketchIt doesn’t have to be exquisite, just use stick figures + sticky notes; (2) Use the real environmentPhoto collage+ Annotation (suitable for field research); (3) Use Figma/whiteboard, etctool generated online

Display method 2: Data charts

(1) Data chart of distribution dimension: Answer “How is the data dispersed?” ”

1. histogram: Displays the distribution of continuous data, and uses a column to represent the frequency of each interval. It can be used to analyze the distribution of user active time and observe the concentrated trend of continuous variables such as age/income → [Histogram of user daily usage time]
2. Density map: Use a smooth curve to show the probability density of the data distribution. Discover potential multi-peak distribution of user behavior patterns (such as morning and evening peak usage) → [APP startup time density distribution]
3. Box line diagram: Displays the minimum, quartiles, maximum, and outlier values of the data. It can be used to compare the discrete consumption amount of different user groups and identify extreme users (such as ultra-high payers) → [Box plot of user payment amount by city]

(2) Data charts that make up dimensions: Answer “What is the proportion of each part?” ”

1. Pie chart/donut chart: Shows the proportional relationship of each part in the whole. It can be used to represent the proportion of user source channels, feature usage frequency distribution, geographical distribution, and transfer distribution (mutual transformation of values over time or conditions)→ [Proportion of new user registration channels]
2. Stacked bar charts: The bar segment shows the proportion of subcategories in the population. It can be used to display the combination of functions preferred by age group, and can also show the cross-composition of user gender and payment status at the same time → [Gender× Payment Tier Stack Chart]
3. Tree diagram: Represents the proportion of hierarchical data with nested rectangular areas. Affiliations that can be used to visualize user interest tags (e.g., “Sports→ Fitness→ Yoga”→ [Interest Classification Tree Chart]

(3) Data chart of the comparison dimension: Answer “Is there a significant difference between groups?” ”

1. Comparison bar chart: Parallel bars compare the numerical differences between different groups. It can be used to compare the conversion rate of scheme A and B in A/B testing, and represents the satisfaction score of different user groups → [NPS comparison bar chart of old and new versions]
2. Radar map: Comparison of multidimensional data on axes, showing relative strength. It can be used to compare the differences of the three types of user groups in 5 experience indicators, or to show the group differences in product function preferences → [User Group Characteristics Radar Chart]
3. Slope chart: Use line segments to connect the values of two time points to highlight the changing trend. It can be used to show the group difference in user retention rate from January to June, or to compare the key indicators before and after the feature revision → [Slope chart of retention changes by level of paid users]

(4) Data chart of correlation dimension: Answer “Is there a correlation between variables?” ”

1. Scatter plot: Use the point position to show the correlation between the two variables, and add a trend line. It can be used to analyze the relationship between the number of active days of a user and the amount of consumption, and find the correlation between the frequency of use and the complaint rate → [Usage time vs. satisfaction scatter plot]
2. Bubble diagram: An upgraded version of the scatter plot, which represents the third dimension with the size of the bubble. It can be used to analyze user age (X), income (Y), consumption frequency (size), or geographical distribution, customer unit price, user size, → [age-income-consumption bubble chart]
3. Sanki Tu: Use flowing lines to show multi-stage conversion or correlation relationships. Can be used to represent the page jump path in the user journey, or the final conversion destination of users across different channels → [Sankey diagram of churn at each step of the sign-up process]
4. Heat map: Indicates the density or intensity of matrix data by color shade. User interest can be analyzed according to the page click density, or display user interest-behavior cross-popularity → [APP homepage click heat map]

Display method 3: User portrait and user journey diagram

User portraitsis a “character script”,User journey mapIt is “the plot unfolds”. User personas are fictional characters built based on real user data, representing a group of users with similar characteristics (behaviors, needs, pain points), mainly answering the question “Who is the user?” The user journey chart visualizes the complete process of user interaction with the product/service, including behavior, emotions, pain points, and opportunity points, mainly answering the question “How does the user experience?” ” The essence of user portraits is:User clustering and difference.Identify nature by mining user basic, psychological, and behavioral attributesclusteringand highlight the key between groupsdifferenceFinally, build a representative user profile to guide product decision-making. The drawing of the user journey map cannot be separated from the specific user type, and it must be bound to a specific user profile.

epilogue

The current user research field is mired in formalism, like academic papers. A large number of reports are written thicker than the Xinhua dictionary, making them “correct nonsense” piled up with data, which neither gives clear conclusions nor is powerless to stop wrong decisions. In the field of user research, there are gradually a thick stack of self-touching 200 pages of unpaid reports, archaeological research that uses outdated data to guide current decision-making, propositional compositions that make conclusions before finding evidence, and zoo-style observations that are out of touch with real scenes. Real user research should dare to break the team illusion, and be able to sniff out business opportunities from users who casually say “can you find a cheaper purchasing agent”, instead of piling up data in PPT all day long, and finally producing an “academic masterpiece” that no one reads except yourself.