Two missing features can provoke very different reactions. One makes customers question whether the product is usable at all; the other passes unnoticed until someone shows them what it could do. The Kano model helps describe that difference, connecting the presence or absence of an attribute with satisfaction.

The model distinguishes basic expectations, performance attributes and attractive extras. Basic attributes are taken for granted until they fail. Performance attributes tend to improve satisfaction as delivery improves. Attractive attributes can please customers without necessarily causing dissatisfaction when absent. Indifferent and reverse responses allow for features that make little difference or are actively unwelcome. Qualtrics’ explanation of Kano analysis describes the paired questions used to investigate these relationships.

Ask about both presence and absence

A Kano questionnaire asks how someone would feel if an attribute were present, then how they would feel if it were absent. The combination of responses informs its classification. This is different from simply asking respondents to rate importance, although importance questions can contribute additional context.

The attribute needs a clear description. “Good security” is too vague to tell people what they are evaluating. A specific capability, such as requiring a second verification step when signing in, gives respondents something more concrete, while still requiring careful explanation of the circumstances. An introductory pilot can reveal whether the paired questions make sense to the intended audience.

In an illustrative collaboration-tool study, automatic meeting summaries might appeal to one group while another worries about sensitive conversations being recorded. Combining their answers into a single label could hide the most consequential finding: the feature means different things to different users.

Use the classification to deepen a decision

Kano categories are not permanent properties of features. Expectations can change as people gain experience and alternatives develop. They also depend on the audience, wording and version of the capability being described. Report the distribution of responses and any meaningful group differences, rather than presenting the most common category as unanimous agreement.

The model does not determine engineering effort, commercial value or an acceptable level of risk. Nor does a basic classification mean further investment is always wasted: reliability and accessibility may need improvement even when customers assume they are already adequate.

Used within feature prioritisation, Kano can help explain why an unglamorous repair deserves attention or why an attractive addition warrants further concept testing. Its contribution is a more nuanced account of satisfaction, which the team must connect to the practical choices ahead.