Many times, product managers will be overwhelmed by the rolling demand. Which users are the needs that users really like? What are some false demands or negative demands? This requires the product manager to consider and distinguish it. Recently, Brian O'Neill, a senior designer of NVIDIA, published an article in Medium introducing a very useful analytical model-Kano Model? )。
in 1984, Noriaki Kano, a professor at Tokyo Institute of Technology in Japan, put forward a model to predict people's satisfaction with a product. Since then, Carnot model has gradually become a standard design tool, because it can effectively make people judge the quality of a function or idea. The core principle of this model is that user satisfaction can be plotted along five distinct curves.
just like necessary functions, there are limits to users' satisfaction. Before a certain point, these functions will not have any "benefits". Pleasant functions are an important part of the user experience and should not be ignored. But they have a long shelf life, in part because they are easy to imitate.
combining all these functions not only provides a clear graphic representation, but also helps you find out the strategic direction.
expected functions: a lot of resources should be invested in these functions, because they are the key to acquiring and retaining users, and they are also competitive advantages.
required functions: before the basic needs are met, a lot of resources should be invested, and after they are met, there is no need to invest a lot of resources.
pleasant functions: it is ok to invest resources here, but don't sacrifice the user's expectations and necessary functions. However, it is worth noting that pleasant functions are often the key to building user loyalty and word of mouth.
irrelevant functions and anti-functions: invest resources in identifying these functions to avoid wasting more resources.
to determine which feature belongs to which category, we need to ask our users. But remember, users are usually not good at identifying or expressing the functions they really want and need. Therefore, Carnot analysis needs to explain a problem by asking questions in pairs: one is an intuitive function problem, followed by a dysfunction problem.
For the above question, users must choose one of the following answers:
I like it
I hope it
I am neutral about it
I can accept it
I don't like it
You can prepare a complete questionnaire for every function in your to-do list. Then, you can analyze each user's answer by drawing the results in the table below.
The analysis table will tell you how users feel about each function, so that you can figure out which function belongs to which category in the Carnot model.
it should be clear that if users like a function when it exists, and they don't like it when it doesn't exist, then this is a function that users expect. When the answers are obviously contradictory, there will be problems. (This usually means that users don't understand these problems.
after you summarize all the results, you can calculate the satisfaction and dissatisfaction coefficients. Satisfaction coefficient is a number between 1 and 1: the closer it is to 1, the greater the influence of this function on users' satisfaction. Dissatisfaction coefficient is a number between 1 and -1: the closer it is to -1, the greater the influence of this function on users' dissatisfaction. We use the following two formulas to calculate the coefficient:
For example, The final findings are as follows:
Expectation: 5%
Necessary: 12%
Pleasant: 4%
Irrelevant: 23%
Anti-function: 31%
Contradictory answer: 25%
So we can draw the following conclusions:
Satisfaction: (4+5+12+23)*(-1)=-1.3864
Carnot analysis is not only low-cost and easy to implement, but also provides clear results for judging the product functions that users really want. It also provides hard data to free everyone from the trap of prejudice or short-sighted thinking. There is no need to waste your breath arguing and discussing with internal stakeholders which functions should be added or excluded. Numbers don't lie!