KANO model is a useful tool to classify and prioritize user needs. Based on the analysis of the influence of user demand on user satisfaction, the nonlinear relationship between product performance and user satisfaction is reflected. In the Carnot model, the quality characteristics of products and services are divided into four types: (1) essential attributes; ⑵ Expected attribute; (3) Charm attribute; (4) Indifferent attributes.

Charm attribute: unexpected by users. If this requirement is not provided, user satisfaction will not be reduced, but when this requirement is provided, user satisfaction will be greatly improved;

Expectation attribute: when this requirement is provided, user satisfaction will be improved, and when this requirement is not provided, user satisfaction will be reduced;

Necessary attributes: when this requirement is optimized, user satisfaction will not be improved, and when this requirement is not provided, user satisfaction will be greatly reduced;

No difference factor: whether this demand is provided or not, user satisfaction will not change, and users simply don't care;

Reverse attribute: users don't have this demand at all, but after providing it, user satisfaction will decline.

KANO questionnaire consists of positive and negative questions for each quality characteristic, which respectively measures the user's reaction when facing the existence or non-existence of a quality characteristic.

In addition to the discussion of Kano attribute attribution, we can also calculate the merit coefficient by classifying the percentage of functional attributes, indicating that a certain function can increase satisfaction or eliminate the degree of influence that we don't like very much.

Increased satisfaction coefficient? Better/SI=(A+O)/(A+O+M+I)

Dissatisfaction coefficient after elimination? Worse /DSI=- 1*(O+M)/(A+O+M+I)

? Better can be explained by the increase of satisfaction coefficient. The value of better is usually positive, which means that if a certain functional attribute is provided, user satisfaction will be improved; The greater the positive value/the closer it is to 1, the greater the impact on user satisfaction, the stronger the impact on the improvement of user satisfaction, and the faster the improvement.

? Worse can be called the dissatisfaction coefficient after elimination. Its value is usually negative, which means that if a function attribute is not provided, the user's satisfaction will be reduced; The more negative/closer the value is to-1, it means that it has the greatest impact on user dissatisfaction, and the stronger the impact of satisfaction reduction, the faster the decline.

Therefore, according to the merit coefficient, priority should be given to the function/service requirements with higher absolute scores.

According to the value of merit coefficient, the scatter plot is divided into four quadrants.

The first quadrant indicates that the value of the good coefficient is higher and the absolute value of the poor coefficient is higher. The attribute falling into this quadrant is called expected attribute, which means that if the product provides this function, the user satisfaction will be improved, and if it does not provide this function, the user satisfaction will be reduced. This is the competitive attribute of quality, and we should try our best to meet the expected needs of users. Provide additional services or product functions that users like, so that their products and services are superior to competitors and different, and guide users to strengthen their goodwill towards the products;

The second quadrant indicates that the value of the good coefficient is higher and the absolute value of the poor coefficient is lower. The attribute falling into this quadrant is called charm attribute, which means that if this function is not provided, user satisfaction will not decrease, but if this function is provided, user satisfaction and loyalty will be greatly improved;

The third quadrant indicates that the value of the good coefficient is low, and the absolute value of the poor coefficient is also low. Attributes that fall into this quadrant are called indifference attributes, that is, whether these functions are provided or not, user satisfaction will not change, and these function points are functions that users don't care about.

The fourth quadrant indicates that the better coefficient is lower and the absolute value of the worse coefficient is higher. Attributes that fall into this quadrant are called required attributes, which means that when the product provides this function, the user satisfaction will not be improved, but when it does not provide this function, the user satisfaction will be greatly reduced; Explain that the functions that fall into this quadrant are the most basic functions, and these requirements are what users think we have an obligation to do.

Among the functions of the same type, it is suggested that the function with higher good coefficient and lower poor coefficient should be given priority.

In product development, the order of function priority is generally: required attribute >; Expected attribute >; Charm attribute > indifference attribute.

However, it is necessary to consider the marketing strategy of the product, such as expectation attribute and charm attribute, which can hit the user's cool or itchy point. Expectation attribute and charm attribute are more important in gaining market share, and packaging marketing can be considered as a selling point of products.

2. Carnot model practice

2. 1 questionnaire writing:

Because KANO model questionnaire needs to know the following two aspects: the user's evaluation (attitude) when the product/service has a certain function and the evaluation (attitude) when the product/service does not have a certain function, it is necessary to ask the user forward and backward respectively. It should be noted that:

? ①? There are positive and negative questions related to each function point in KANO questionnaire, so it is necessary to correct the difference between rhetorical questions to prevent users from misreading the meaning of the questions.

? ②? In the actual topic setting, when the number of function points is relatively large (more than 5) or there are little differences and similarities between function points, it is suggested that users should be grouped, and each user should answer at most 5 function points, so as to make a function point with high discrimination as far as possible.

? ③? In the question type, it is suggested to give priority to single-choice questions and avoid using array questions, because under array questions, users are more likely to answer randomly or indiscriminately, resulting in not distinguishing each function point, such as all expected functions.

? ④? Function description: briefly describe the function points to ensure users' understanding;

? ⑤? Description of options: Because users have different understandings of "I like it very much", "I should", "I don't care", "I reluctantly accept it" and "I don't like it very much", it is necessary to give a unified explanation before filling out the questionnaire, so that users can have a relatively consistent standard and answer it conveniently.

I like it very much: it makes you feel satisfied, happy and surprised.

It should be: a function/service that you think is necessary.

It doesn't matter: you don't particularly care, but it is acceptable.

Reluctantly accept: you don't like it, but you can accept it.

I don't like it: it makes you feel dissatisfied.

? ⑥? Add criteria to verify KANO results or crowd analysis: the frequency of users' use of functions (background data, questionnaire query), which functions users like best (if each function point has a small degree of discrimination, it can be re-divided from the degree of preference if it belongs to the desired attribute), and which functions users will choose to use products for.

2.2 data analysis

Data cleaning → Carnot two-dimensional attribute attribution analysis → merit coefficient calculation. It can be analyzed directly in Excel or SPSS.

In addition, it can also be combined with some data of the product to support combinatorial analysis, such as user portrait, UV, conversion rate and so on.

2.3 data interpretation

KANO mode is to discuss the priority of functions/services, and the specific situation needs to be discussed with business parties. The results of Kano model are combined with the actual business situation to determine the priority of feasible product function development/optimization, so as to implement the research results.