Kano Classes
The Kano model sorts requirements by how fulfilling them changes satisfaction, and the answer is not the same for all of them. Three classes matter [1]: must-be, one-dimensional and attractive. The class is a property of a customer segment, not of the feature.
This answers the question the satisfaction ratio leaves open: if satisfaction is performance over expectation, why did delivering everything they asked for still disappoint them?
1. The three classes
- Must-be. “If these requirements are not fulfilled, the customer will be extremely dissatisfied. On the other hand, as the customer takes these requirements for granted, their fulfillment will not increase his satisfaction.” In software: the login works, the data is not lost, the page loads. These can only lose you points.
- One-dimensional. Satisfaction is “proportional to the level of fulfillment” — a straight line. Speed, storage quota, API rate limit. This is usually what the customer asks for.
- Attractive. “Neither explicitly expressed nor expected by the customer. Fulfilling these requirements leads to more than proportional satisfaction. If they are not met, however, there is no feeling of dissatisfaction.” These can only win points.
The software readings are ours; the source’s examples are all skis.
Industry names above, Sauerwein’s below. The migration arrows are not in the 1996 paper. [1]
There is a tie-break rule, and it is the practical payload: must-be beats one-dimensional beats attractive beats indifferent. Failing a must-be cannot be compensated by any amount of delight.
2. How a class is determined
Not by asking how important something is. Kano uses a question pair for each requirement — a functional question (how would you feel if the product had this?) and a dysfunctional one (how would you feel if it did not?), each answered on a five-point scale, then read through a classification matrix. One question cannot produce a category, which is exactly why an importance-weighted backlog cannot either.
The method yields two satisfaction coefficients per requirement — one for how much fulfilling it raises satisfaction, one for how much omitting it lowers satisfaction. For example, in the study’s own worked case ski edge grip scored +0.40 / −0.83 while ski servicing scored +0.89 / −0.25. Read those as investment cases rather than as ratings: the first is insurance — getting it right buys little, getting it wrong is punished hard — and the second is the opposite. A single importance score cannot distinguish them, and that is the failure mode this model exists to fix.
3. Two boundary conditions that carry as much weight as the model
The category belongs to the segment. In the study, edge grip was must-be to expert skiers and one-dimensional to novices. The same is true of a software feature: single sign-on is a must-be for your enterprise accounts and an irrelevance to your free tier. A Kano study run over an unsegmented population produces averages describing nobody.
Attractive decays into must-be. A phone camera delighted in 2007; its absence is disqualifying now. Free HTTPS, dark mode and single sign-on have all made the same journey. So a classification has a half-life, and a roadmap built on a three-year-old one is defending a position that has already moved. (That example is ours, not the study’s.)
There is also a cost ceiling worth knowing before commissioning the work: the study reports that 20 to 30 interviews per homogeneous segment reach 90–95% of the requirements. Elicitation has a knee, and interviewing past it is waste.
4. When not to use it
Kano needs a segment you can name and customers you can ask a paired question. It has nothing to offer where the requirement is non-negotiable — a legal or safety obligation is a must-be by definition, and classifying it wastes the study. It also says nothing about cost, so a Kano class is an input to prioritisation and never a prioritisation on its own: an attractive requirement costing three months loses to a one-dimensional one costing three days.
How solid is this?
- Where it comes from. One 1996 study of ski equipment, with more than 1,500 respondents and a questionable-response rate under 2%, presented at a production-economics working seminar.
- What is contested. The paper does not bound its own method, and later work on the model’s reliability is not held here.
- What we do not hold. No replication on a software product in this bibliography. The model transfers as a way of thinking; the categories and coefficients do not transfer from skis.
References
- E. Sauerwein, F. Bailom, K. Matzler, and H. H. Hinterhuber,“The Kano Model: How to Delight Your Customers,” in Preprints Volume I of the IX International Working Seminar on Production Economics, Feb. 1996, pp. 313–327.
Disclaimer: AI is used for text summarization, polishing and explaining. Authors have verified all facts and claims. In case of an error, feel free to file an issue.