The Kano model: classifying product features by their impact on satisfaction
The Kano model divides product features into categories according to how their presence and absence affect customer satisfaction. The article describes Noriaki Kano's five categories, the questionnaire with a pair of questions, a worked example classifying three features and the method's limitations.
Amounts and values in the examples are illustrative and show how the calculation works. In projects, we use the client's data.
Dominik Dębski · Lemur Media · · reading time: 4 min
Reviewed by Dominik Dębski, 1 November 2026
What is the Kano model?
The Kano model classifies product features by their asymmetric impact on satisfaction: the absence of some features causes strong dissatisfaction, while the presence of others creates satisfaction the customer did not expect. The model was described by Noriaki Kano, Nobuhiko Seraku, Fumio Takahashi and Shinichi Tsuji in the article "Attractive quality and must-be quality", published in 1984 in the journal of the Japanese Society for Quality Control.
What are the feature categories?
| Category | Feature present | Feature absent | B2B SaaS example |
|---|---|---|---|
| Must-be | Does not increase satisfaction | Strong dissatisfaction | Data export, backups |
| One-dimensional | Satisfaction rises with the level of the feature | Dissatisfaction | Speed, number of integrations |
| Attractive | Satisfaction the customer did not expect | No dissatisfaction | Automatic AI-generated summaries |
| Indifferent | No effect | No effect | Choice of colour theme |
| Reverse | Dissatisfaction among some customers | Satisfaction among some customers | Extensive notifications |
Categories change over time. An attractive feature becomes one-dimensional and then must-be once competitors introduce it and customers start to expect it.
How does the survey work?
For each feature, the respondent answers a pair of questions: a functional one about a situation in which the product has the feature, and a dysfunctional one about a situation in which it does not. Answers are chosen from five options: I like it, I expect it, I am neutral, I can tolerate it, I dislike it. An example pair for a project management system: rating a situation in which the system itself reminds users of upcoming deadlines, and rating a situation without such reminders. The questions describe the benefit to the user, without technical names. The combination of the two answers determines the category:
- like it when present and dislike it when absent: one-dimensional,
- like it when present and neutral when absent: attractive,
- expect it when present and dislike it when absent: must-be,
- neutral on both questions: indifferent,
- dislike it when present and like it when absent: reverse.
What does a worked example look like?
Each feature is assigned the category chosen by the largest number of respondents, and close results for two categories are analysed by segment. The example covers 100 respondents from customer companies of a project management system.
| Feature | Attractive | One-dimensional | Must-be | Indifferent | Reverse | Category |
|---|---|---|---|---|---|---|
| Automatic deadline reminders | 18 | 52 | 15 | 13 | 2 | One-dimensional |
| Report export to PDF | 5 | 15 | 62 | 17 | 1 | Must-be |
| AI planning assistant | 41 | 12 | 4 | 35 | 8 | Attractive |
PDF export must work flawlessly, but expanding this feature will not increase satisfaction. Reminders are worth developing, because satisfaction rises with their quality. The AI assistant is attractive to 41% of respondents but indifferent to 35%, which points to two segments with different needs; before an investment decision, the answers are analysed by company size and respondent role.
How do you link the results to roadmap planning?
Kano categories translate into planning rules: must-be features must work reliably, one-dimensional features are developed to the level of the competition or above, and attractive features are tested as differentiators.
| Category | Planning rule | Success measure |
|---|---|---|
| Must-be | Reliability level without expanding scope | Number of bug reports and complaints |
| One-dimensional | Gradual improvement benchmarked against competitors | Survey rating and impact on retention |
| Attractive | Limited test with a selected group of customers | Feature usage and impact on won bids |
| Indifferent | No investment | n/a |
| Reverse | Option to switch off or off by default | Number of switch-offs |
Kano results are worth comparing with an effort estimate. A low-cost attractive feature goes into the next quarter, while an expensive attractive feature first needs a prototype and confirmation in the segment where it has the most supporters.
What are the model's limitations?
- The survey measures stated preferences; features that customers cannot imagine require a prototype or a description.
- A questionnaire with two questions per feature limits the number of features in one survey to a dozen or so.
- The model does not account for the cost of building a feature; the result is combined with an assessment of cost and revenue impact.
- Different customer segments may classify the same feature differently, so a result for the whole sample without segmentation can be misleading.
- The classification ages with the market and needs repeating, e.g. every year.
Sources
- Wikipedia: Kano model
- J-STAGE: Noriaki Kano, Nobuhiko Seraku, Fumio Takahashi, Shin-ichi Tsuji, Attractive Quality and Must-Be Quality, Journal of the Japanese Society for Quality Control 14(2), 1984
- The Center for Quality Management Journal: Kano's Methods for Understanding Customer-defined Quality, vol. 2, no. 4 (1993)