Demand analysis method-KANO model

1. Overview

1.1 Background

As a product manager, Li Lei often encounters a lot of product requirements. The development classmates are too busy to deal with, and the users seem to want everything. With limited product development resources, how can we find out the real user needs? Give high priority to really important needs?

Li Lei decided to introduce the "KANO model" to sort out the needs of the system, analyze and refine the needs, and improve efficiency.

1.2 Analysis process

KANO model: It is a useful tool to classify and prioritize user needs. It is based on analyzing the impact of user needs on user satisfaction and reflects the nonlinear relationship between product performance and user satisfaction.

Li Lei conducted a survey on about 100 users and used the KANO model to draw a four-quadrant diagram, as shown in the following figure:

The four quadrants correspond to the four types of needs, and their priorities are: Must-be demand>One-dimensional demand>Attractive demand>Indifferent demand.

• Must-be demand (must have): the pain points that are often said. For users, these needs must be met, of course. When this demand is not provided, user satisfaction will be greatly reduced. This category is the core requirement and the required function of the product.

• One-dimensional demand (should have): When this demand is provided, user satisfaction will increase; when this demand is not provided, user satisfaction will decrease. Usually used as the focus of comparison between competing products.

• Attractive demand (may have): Surprise product features that exceed user expectations and often bring high loyalty. Not providing it will not reduce user satisfaction.

• Indifferent demand (optional): Needs that users don't care about at all have no impact on user experience. Try to avoid doing this type of function.

1.3 Analysis results

Li Lei shared the dashboard he made with his colleagues, and decided to add "Function 2, Function 3, Function 5, Function 8" to this function update. With the support of data, everyone agrees with his decision. It is rare that there has been no overwhelming battle over which function to add in the past, and efficiency has been improved.

1.4 Get the dashboard

Click to view the dashboard: KANO template. After saving the dashboard, the user can view the detailed operations in the learning dashboard.

2. Implementation

2.1 Survey questionnaire design

Each function in the KANO questionnaire must have forward and reverse questions, for example:

Clean the data after the investigation. Download the cleaned data in this case: KANO raw data.xlsx

2.2 Process data

1) Upload "KANO raw data" to FineBI. Add a self-service dataset and check all the fields of "KANO raw data", as shown in the figure below:

2) Add new column "Merge Attitude", merge "Add Function Attitude" and "Not Add Function Attitude", as shown in the figure below:

According to the user's "Add Function Attitude" and "Not Add Function Attitude", we can finally use the following table to locate what a fnction is for users.

M: Must-be /basic demand; O: One-dimensional/performance demand; A: Attractive/excitement demand;

I: Indifferent demand; R: Reverse demand; Q: Questionable result.

3) In the last step, we already know how to locate the demand type, the next thing to do is to locate the judgment in the analysis table and add the type column, as shown in the following figure:

Since the formula is very long, the user can directly copy the following to the formula column, and replace the "Merge Attitude" with their own fields, using the switch function:

SWITCH (Merger Attitude, "like very much like very much","Q","like very much should be so","A","like very much it doesn't matter","A","like very much reluctantly accept","A","like very much dislike very much","O","should be so like very much","R","should be so should be so","I","should be so it doesn't matter","I","should be so reluctantly accept","I","should be so dislike very much","M","it doesn't matter like very much","R","it doesn't matter should be so","I","it doesn't matter it doesn't matter","I","it doesn't matter reluctantly accept","I","it doesn't matter dislike very much","M","reluctantly accept like very much","R","reluctantly accept should be so","I","reluctantly accept it doesn't matter","I","reluctantly accept reluctantly accept","I","reluctantly accept dislike very much","M","dislike very much like very much","R","dislike very much should be so","R","dislike very much it doesn't matter","R","dislike very much reluctantly accept","R","dislike very much dislike very much","Q")

Results as shown below:

4) Add "Group Summary" to get the number of people with various demand types for each function, as shown in the following figure.

For example, among the number of people participating in the survey, 48 people think that "function1" has no different needs.

5) Because some users skip questions during the survey process, the number of people participating in each function survey is different. Add a new column "Number of people participating", select "within the group all values", as shown in the figure below, in order to find the number of people participating in each function survey.

6) Calculate the proportion and find the proportion of each demand type in the number of people participating in the survey.

For example, the proportion of type "I" people of "function1" to the number of people participating in the "function1" survey is 0.48, as shown below:

2.3 Make the component

1) Copy 5 "proportion" fields, as shown in the figure below:

2) Perform detail filtering on the copied "proportion1" field, and the filter condition is: the "type" is in "A". And rename it to "A proportion", as shown below:

In the same way, perform detail filtering on the other copied "proportion" fields, respectively filter the types, and rename them, as shown in the following figure:

3) Use the better-worse coefficient, as shown in the figure below.

• better-Satisfaction coefficient improved after adding a certain function: better=(A proportion + O proportion)/(A proportion + O proportion + M proportion + I proportion), the closer to 1, the user satisfaction The stronger the effect of improvement, the faster the increase in satisfaction.

• worse-Dissatisfaction coefficient if not adding a certain function: worst=-1*(O proportion +M proportion)/(A proportion +O proportion +M proportion +I proportion), the closer it is to -1, It means that it has the greatest impact on user dissatisfaction. The stronger the effect of reducing satisfaction, the faster the decline.

According to the above formulas for "better" and "worst", create a new calculation field "better" and "worse absolute value", as shown in the following figure:

4) Select "Scatter Chart" and drag in the "better" and "worse absolute value" fields. And drag the "function" field into the label bar and color bar of the "Graphic Properties", as shown below:

5) Add "cordon" and "vertical warning line" respectively, which are better average and worse average respectively, as shown below:

2.4 Effect display

The effect is shown in section 1.2.