5.4 The Kano Model and Quality Function Deployment
Key Takeaways
- Kano categories are must-be, one-dimensional, attractive, indifferent, and reverse.
- Must-be requirements cause dissatisfaction when absent but produce no satisfaction when present, so improving them beyond the threshold wastes resources.
- Attractive requirements decay over time into one-dimensional and eventually must-be requirements as customer expectations rise.
- The house of quality relates weighted customer whats to technical hows, and the roof matrix records correlations among the technical requirements.
- A technical importance score is the sum over customer requirements of customer weight times relationship strength.
The Kano model
Noriaki Kano's insight is that the relationship between how well a requirement is met and how satisfied the customer feels is not the same for every requirement. Five categories result.
| Category | Behaviour | Example | Project implication |
|---|---|---|---|
| Must-be (basic, dissatisfier) | Absence causes strong dissatisfaction; presence produces no satisfaction | Brakes work; the invoice arithmetic is right | Meet the threshold reliably; do not over-invest beyond it |
| One-dimensional (performance, satisfier) | Satisfaction rises roughly linearly with performance | Fuel economy; delivery lead time | This is where competitive differentiation and most DMAIC value lives |
| Attractive (delighter, exciter) | Absence causes no dissatisfaction; presence produces disproportionate satisfaction | Unexpected proactive status update | Source of differentiation; usually a DFSS rather than DMAIC target |
| Indifferent | Customer does not care either way | Internal documentation format | Candidate for cost reduction |
| Reverse | More of it makes the customer less satisfied | Excessive feature complexity; over-frequent notifications | Reduce deliberately |
Two consequences dominate the exam:
- Over-investing in must-be requirements produces no satisfaction gain. Taking on-time delivery from 99.2% to 99.6% in a market where anything above 99% is assumed buys nothing. The same effort spent on a one-dimensional attribute would move the customer.
- Failing a must-be requirement cannot be compensated by excellence elsewhere. A delightful interface on a system that loses transactions is an unhappy customer.
Attribute decay
Categories are not permanent. An attractive attribute becomes one-dimensional as competitors copy it, then must-be as customers come to assume it. Wi-Fi in hotels moved through all three in about a decade. The implication for a Black Belt is that Kano classification has a shelf life and should be refreshed rather than inherited from an old study.
Classifying requirements
Kano classification uses a paired-question survey. For each attribute, ask the functional form ("how would you feel if the order arrived within 24 hours?") and the dysfunctional form ("how would you feel if it did not?"). Both use the same five-point scale: I like it, I expect it, I am neutral, I can tolerate it, I dislike it. The pair of answers maps to a category through a standard evaluation table -- for example, like on the functional question plus dislike on the dysfunctional question indicates one-dimensional, while expect plus dislike indicates must-be.
Quality function deployment
QFD is the structured method for translating the voice of the customer into technical requirements and carrying that translation through design. Its central artifact is the house of quality.
The rooms
| Room | Contents | Question answered |
|---|---|---|
| Left wall | Customer requirements (the whats) with importance weights | What does the customer want? |
| Ceiling | Technical requirements or design characteristics (the hows) | What can we measure and control? |
| Body / relationship matrix | Strength of each what-how relationship, typically 9 / 3 / 1 | Which technical characteristics affect which needs? |
| Right wall | Competitive assessment from the customer's perspective | How do we compare on each need? |
| Roof | Correlations among technical requirements | Which design choices conflict with each other? |
| Basement | Technical importance scores, targets, and technical benchmarks | Where do we set the specifications? |
Computing technical importance
For technical characteristic $j$:
where $w_i$ is the importance weight of customer requirement $i$ and $r_{ij}$ is the relationship strength between requirement $i$ and characteristic $j$ (9 strong, 3 moderate, 1 weak, blank none).
Worked example for a delivery service:
| Customer requirement (what) | Weight | Route density | Dispatch cycle time | Vehicle reliability |
|---|---|---|---|---|
| Arrives when promised | 5 | 3 | 9 | 9 |
| Short lead time | 4 | 9 | 9 | 1 |
| Undamaged goods | 3 | 1 | 0 | 3 |
| Technical importance | $5(3)+4(9)+3(1) = 54$ | $5(9)+4(9) = 81$ | $5(9)+4(1)+3(3) = 58$ |
Dispatch cycle time carries the highest technical importance, so it is where specification effort and project resources should go first.
Reading the roof
The roof records whether improving one technical characteristic helps or hurts another: strong positive, positive, negative, strong negative. A negative correlation is a design conflict that must be resolved by trade-off or by innovation rather than by wishing. Increasing route density improves lead time but may reduce vehicle reliability through higher duty cycles -- that conflict is exactly what the roof is for.
The four-phase cascade
QFD does not stop at one house. The outputs of each house become the inputs of the next:
- House 1: customer requirements to technical requirements.
- House 2: technical requirements to part or component characteristics.
- House 3: part characteristics to process parameters.
- House 4: process parameters to production and control requirements.
House 4 is where QFD connects to the Control phase, because production requirements become control plan entries. This cascade is what makes the link from a customer statement to a control chart on a specific process parameter traceable.
Using Kano and QFD together
Kano supplies the weights for the left wall. Requirements are not equally important, and Kano category tells you how importance behaves:
- Must-be requirements get a threshold target rather than a maximization target. Meet them and stop.
- One-dimensional requirements carry the highest improvement weight, because satisfaction scales with performance.
- Attractive requirements are weighted for differentiation value but should not displace unmet must-be requirements.
- Indifferent requirements are candidates for cost removal.
A house of quality built with unweighted customer requirements, or with weights that ignore Kano category, will systematically over-specify basics and under-specify differentiators.
A hotel chain currently achieves 99.2% availability of hot water, an attribute that guests treat as assumed. A project proposes to raise it to 99.7%. What does the Kano model predict?
In a house of quality, customer requirements are weighted 5, 4, and 3, and a technical characteristic has relationship strengths of 9, 9, and 0 with them respectively. What is its technical importance score?
What does a strong negative entry in the roof of the house of quality indicate?