Kano Model

The Kano model classifies product attributes by how their presence or absence affects customer satisfaction. Developed by Noriaki Kano in the 1980s, it distinguishes categories that behave differently: must-be attributes, which cause dissatisfaction when absent but no satisfaction when present; one-dimensional attributes, where satisfaction rises in proportion to performance; and attractive attributes, which delight when present but are not missed when absent. Indifferent and reverse categories cover attributes customers do not care about and those that reduce satisfaction by being present.

The model's contribution is showing that features are not interchangeable units of value. Must-be attributes, such as a site loading, a payment working, or an order arriving, generate no goodwill when delivered and severe dissatisfaction when they fail, which means investment in them is defensive rather than differentiating. Attractive attributes generate disproportionate positive response for the same effort, but only once the must-be attributes are satisfied. A product that delights in novel ways while failing at basics accumulates dissatisfaction regardless of how impressive the additions are.

The most practically important element is the model's temporal dimension. Attributes migrate over time: what was attractive becomes one-dimensional and eventually must-be, as customers come to expect it and competitors adopt it. Free delivery, order tracking, mobile responsiveness, and same-day support have each followed this path. The implication is that a product standing still becomes progressively less satisfying without changing at all, and that competitive advantage from a delighter is temporary by construction.

The model comes with a specific survey instrument, in which each attribute is presented twice, functionally and dysfunctionally, asking how the respondent would feel if the feature were present and if it were absent. The pattern of paired responses classifies the attribute. The questionnaire is more demanding than it looks: it is long when many attributes are tested, respondents find the dysfunctional phrasing confusing, and results are sensitive to how attributes are described. Testing a small number of well-defined attributes with a properly recruited sample produces far better results than a comprehensive survey answered carelessly.

Its main limitations are that classifications differ by segment, so an attribute that is attractive to one customer type may be indifferent or must-be to another, and that the model addresses satisfaction rather than willingness to pay or commercial value. An attractive feature that delights customers may still be unprofitable to build and maintain, and the model provides no guidance on that trade-off.

A lighter-weight application that avoids the survey overhead is to classify a proposed backlog qualitatively during prioritization, asking of each item whether its absence would cause complaints and whether its presence would be noticed. This is less rigorous than the formal instrument but sufficient for the most common practical purpose, which is identifying that a set of proposed enhancements are all in the must-be category and will therefore not differentiate anything. It also surfaces the reverse category, features that some customers actively dislike, which is easy to overlook when a proposal is championed by an enthusiastic internal sponsor and never tested against the wider customer base.

Used within those limits, the framework is a useful corrective to prioritization methods that treat all features as equivalent candidates for the same scoring exercise. In practice it informs roadmap discussions within product research, with the attribute definitions and segment analysis drawn from user research, and its most common practical use in commercial work is establishing that a proposed set of enhancements sits in the must-be category, where the correct expectation is avoiding loss rather than producing gain.