Kano Model

[lior_hertz]

Hi,
I would be happy to get comments about the Kano Model, better from experienced.
What are the advantages and the disadvantages of the model, or what are the cases the usage is recommended?
Thanks,
Lior

[Sherman Whipple]

Lior,

I am not familiar or experineced with a Kano "model" so I can not comment on advantages, disadvantages or recommended usages of such a model.

I am, however, very familair with the Kano test as well as the Kano method using a 4x4 matrix. This test is used to gauge and or track categorical response to the presence or absence of product attribures or features. For reliable results, respondents must be pre-screened for a level of product experience or knowldge.

In my experience, when properly applied, the Kano method is a nearly perfect predictor of consumer/buyer expectations or purchase criteria at the given point of the products evolution for which it was measured.

This, however, does not apply to the ordinal variants of Kano which use a 5x5 matrix and are of only questionable value.

Sherman

[lior_hertz]

Sherman, Thank you very much!
To be more specific- in your experience, does the test work well with services? Or mainly test products?
Could you please explain a bit more about the differences between 5x5 and 4x4 matrix?
For better understanding- can you give an example of one or two questions which used in the test?
Thank you,
Lior

[Sherman Whipple]

Yes, Kano can be used with services, however, it might not be its best use but this would depend on the specifis of the service offering and whether it were a standardized service such as a lube shop versus a customized service.

In a 4x4 Kano, the possible responses are I would be Delighted, I would Expect it to be so, I am Ambivalent, and I would be Disapointed. These are catgories and there is no option for lack of knowlege, or a neutral response or an easy out, sicne it is assumed that the respondent in knowlegable and therefore, for an informed subject, one category must apply both in the case of the presence if the attribute and its absence. Under the scenario of a new luxury car purchase how would you respond to the presence of a satellite radio. How would you respond to its absence. Alternatively under the scenaruio of an economy car, how would you respond to the presence of a satellite radio, how would you respond to the absence?

What the 5x5 variation did was allow researchers to offer a sampling that looked a lot like kano, but let them be sloppy both in the subjects they selected as well as in the analysis of the data by treating the response as ordianl and adding a neutral response as a catch all for those who knew nothing of the subject about which they were being tested, as well as an easy out. Basically it contaminates the table and produces misleading results.

[lior_hertz]

Sherman- thanks again.
I was asking about applying the Kano test in services, but I have a problem which I don’t know how to solve.
For example, “waiting time”, which customers should wait until the representative is answers. How can I ask about this attribute? It’s not as you wrote “a presence or absenity of an attribute”. So if “waiting time” is one of other attributes which a company is thinking to offer in its customer service, how should it be applied it in the Kano test?
Thanks again,
Lior

[Sherman Whipple]

Waiting time is more likely to be a qualitative factor than an attribute and, therefore, not something which could be tested using the Kano method. Rather you would use conjoint or CBC which can be done in conjuction with Kano.

This said, an example of where time can be a Kano testable attribute would be for a glucose meter's result which went from 30 seconds, to 15, and to 5. Six years ago the presence of a 5 second result time largely scored a delighted, and the absense was ambivalent. 2/3 of the meters at thqt time were greater than 15. Within three years, the presence of a 5 second result attribute became a must be, and its absence a disappointment.

[Seán]

My undestanding (and use) of Kano was that it is a theory rather than a specific method. Given interpretation in Kano is about the utility within a context. Wouldn't conjoint analysis with interactions provide a test and then run the simulations to give you the satisfaction curves?