The categories are then related to the product attributes. Oliver (1997)
distinguishes between: bivalent satisfiers—the upward and downward
translatable attributes that can cause both satisfaction and dissatisfaction
(proportional quality in Kano’s model); monovalent dissatisfiers—
essential but unprocessed attributes only capable of causing dissatisfac-
tion when flawed (must-be quality in Kano’s model); and monovalent
satisfiers—psychological extras processed at a higher level of the needs
hierarchy (exiting quality in Kano’s model).
The Kano approach illustrates that it is difficult to achieve customer
satisfaction and his approach covers aspects of both compensatory and
noncompensatory models. So far, the Kano approach has not been mod-
eled and the general second-order model (e.g., used by Sloof et al., 1996)
is not sufficient either.
Identification of Consumer Needs
In general, consumer needs should be identified by means of quali-
tative surveys. In QFD terminology, this is called listen to the customer’s
voice.Various methods can be used, but the most common are personal
interviews and focus groups (Griffin and Hauser, 1993). Laddering in-
terviews can also be used (Oliver, 1997), but usually result in an iden-
tification of needs at higher levels in the hierarchy (Griffin and Hauser,
1993) because they only include statements mentioned by a certain num-
ber of respondents.
For QFD purposes, needs are identified directly from consumers’
statements. It is difficult to say precisely how many interviews are nec-
essary to identify consumer needs satisfactorily, and only one study has
been carried out in the area. Griffin and Hauser (1993) show that about
two-thirds of consumer needs can be identified from four personal in-
terviews or two focus groups, and that this rises to 80% with two extra
personal interviews or one focus group. The value of additional inter-
views declines, and their study does not show an expected effect of group
dynamics in the focus groups. They recommend 20–30 personal inter-
views, while others are more realistic and recommend 10–20 (Day,
1993). According to Griffin and Hauser (1993), it is important that the
results are analyzed by persons with different backgrounds.
Commenting on Kano’s model, Day (1993) says that basic needs
(must-be quality) are rarely mentioned in qualitative studies; it is the
functional attributes that are expected and usually fulfilled. Unless the
customer has experienced the absence of these attributes, they are not
mentioned. Instead, proportional quality is what consumers talk about.