© 2002 QFD Institut Deutschland e.V.
QFD (Quality Function Deployment) tries to understand the role of a product for its customer - to make the product a success in the market place.
Identifying relevant customer requirements is the main problem in QFD projects. The most practical approach is visiting the gemba: the gemba is that place and time, where and when a customer uses the product. Observing the interaction between the customers and his/her product reveals relevant requirements (1).
However, this cannot always be done- think of a new generation product! An alternative then is inventing the gemba by TRIZ.
"Listen to the unspoken" is Glenn Mazur's ongoing theme. "Doing nothing wrong is not: doing anything right!" The problem is: what your customer does not say is important. Noriaki Kano visualized this by a simple model (Figure 1).
Figure 1: Kano Model - Supplier's Ability vs. Customer's Perception
Expected requirements are "self-evident" and unspoken. E.g. one can expect that a car will drive and can stop. Even a 100% fulfillment of these functions will never satisfy a customer - it is expected anyway. On the contrary, when these functions are absent or fail, customer will be dissatisfied. Kano calls them dissatisfiers (2).
Revealed requirements are expressed somehow.
They are known; the degree of fulfillment correlates with customer
satisfaction - like asking for a specific interior of a car. Kano calls
them satisfiers.
Exciting requirements are not expected and are not asked for -
they are unspoken. However, should they be available in the product,
customer can become very excited - like having a GPS-system at no
extra cost. Kano calls them delighters.
Cohen's examples make it easy to identify TRIZ concepts, which are characteristic for dissatisfiers, satisfiers and delighters.
"Dissatisfiers are the absence of 'expected quality', in the sense that customers expect products to be essentially flawless, and if they are not, the customers are dissatisfied. Examples of dissatisfiers are scratches or blemishes on the surface of a product, broken parts, missing instruction booklets, or missing features that are routinely supplied in similar products. Customers don't tell us they want 'expected quality' because they take for granted we will provide it."
From a TRIZ-perspective we have to ask:
A satisfier is something that customers want in their products, and usually ask for. The more we provide a satisfier, the happier customers will be. ( ) Examples of satisfiers are increased capacity, lower cost, higher reliability ( ) In other words, you can expect the satisfiers to be present in all the competitive products, to a greater or lesser extent."
Very clearly from a TRIZ-perspective we should understand:
"Delighters are product attributes or features that are pleasant surprises to customers when they first encounter them. ( ) A typical customer reaction to a delighter is to say to a friend, 'Hey, take a look at this!' ( ) The 3M Post-it Note is another example of a delighter. It's a product that filled user needs that had not been filled satisfactorily. ( ) The needs that delighters fill are often called 'latent' or 'hidden' needs. ( ) These needs are sometimes intimately linked to customer's perceptions of the limits of technology ( )"
Delighters answer these questions:
Why is a customer using a specific product? - The customer has a specific objective (Figure 2); he/she utilizes a specific function from a selected product: when the product performs this function, the customer can achieve his/her expected objective (satisfier). All product functions, which hinder achieving this objective, dissatisfy the customer (dissatisfier). - We can describe delighters in a similar way - when we focus on unexpected value.
It is just like solving a puzzle: frequently, missing information about the customer can be concluded while building this model.
Figure 2: Showing the Kano-Qualities in a TRIZ-Model
Once we found adequate TRIZ-models we can improve the situation, e.g. by invoking the directions from the IWB software (3). For the generic model of Figure 2 we can develop ideas - focused on the customer (Table 1).
Table 1: How to improve the Situation modeled by Figure 2
1. Find an alternative way to obtain [the] (expected useful action /satisfier/) that provides or enhances [the] (known objective). |
2. Find an alternative way to obtain [the] (known objective) that offers the following: provides or enhances [the] (expected value), does not require [the] (expected useful action /satisfier/), is not influenced by [the] (harmful action /dissatisfier/), (missing useful action /dissatisfier/) and (limits or conflicts /hidden needs/). |
3. Find an alternative way to obtain [the] (expected value) that does not require [the] (known objective). |
4. Consider transitioning to the next generation of the system that will provide [the] (expected value) in a more effective way and/or will be free of existing problems. |
5. Find a way to eliminate, reduce, or prevent [the] (harmful action /dissatisfier/). |
6. Find a way to eliminate, reduce, or prevent [the] (missing useful action /dissatisfier/). |
7. Find a way to eliminate, reduce, or prevent [the] (limits or conflicts /hidden needs/). |
8. Find an alternative way to obtain [the] (enhanced useful action /delighter/) that provides or enhances [the] (more than expected). |
9. Find an alternative way to obtain [the] (more than expected) that offers the following: provides or enhances [the] (unexpected value), does not require [the] (enhanced useful action /delighter/). |
10. Find an alternative way to obtain [the] (unexpected value) that does not require [the] (more than expected). |
11. Consider transitioning to the next generation of the system that will provide [the] (unexpected value) in a more effective way and/or will be free of existing problems. |
Utilizing these directions (Table 1) is good for QFD. We can use them to purify customer requirements: "We can implement this idea - is this really, what you want?" Or - we can use them for the Pugh-concept selection chart later in the QFD-process.
"Will Jan wake up? He pushed the button and ink rattled on paper -
each program has its price, hasn't it?"
The market segments of printers include
people, who use printouts for checking their implementation of programs - such as
the author. How to improve this market segment?
The author used different desktop printers. When starting at Philips Semiconductors the LA50 printer was widely used as a shared printer - inadequate for private purposes (Table 2). In the beginning 90's the author needed frequent printouts from source codes; he wanted a private low-noise printer - so he bought the NEC 24P. In the mid 90's he upgraded to a LED-printer from Panasonic, which was faster, less noisy and had a much better image quality. He switched to the inkjet, when hp introduced this trade-off solution in the late 90's.
Table 2: Perceived Advantages and Disadvantages of Printers
Printer | Advantages and Disadvantages |
LA50 (9 needles) | very loud , very slow , very bad printout , free |
NEC 24P (24 needles) | pretty fast , more silent , expensive |
Panasonic (LED) | very fast , very good print , silent , ozone , price |
Hp-Inkjet (ink) | slower , print , color , moderate price , no ozone , no odor |
Table 3: Estimating Perceived Ideality from Table 2
Printer | Perceived Ideality | ||
LA50 | 1 | 3 | 1/3 = 0.33 |
NEC24P | +2 | +1 | 3/4 = 0.75 |
Panasonic | +3 | +2 | 6/6 = 1.00 |
HP-Inkjet | +4 | +2 | 10/8 = 1.25 |
Figure 3: Trend in Perceived Ideality (Table 2, Table 3)
We can estimate the trend in ideality by comparing the new advantages and disadvantages from generation to generation (Table 3, Figure 3). It is a perceived ideality - specific to this market segment. How to identify the Kano-qualities?
Let's focus on the Panasonic-situation. The author wants to review source code - objective.
He buys this printer, because it offers faster and high quality printing; this product function allows
him to achieve his objective - satisfier. Meeting his objective creates value: to create a program with
fewer bugs - expected value (Figure 4).
At home, an unwanted function reveals itself: the printer emits ozone! Tolerated to some
degree, it finally limits the frequency of printouts - dissatisfier.
At home, a nice function revealed itself: the printer is very noiseless! This delighter allows starting
more complex projects; creating programs with more functionality may come true - unexpected value.
Though - the ozone imposes a limit here as well.
Figure 4:TRIZ Model for the Printer
The TRIZ model describes perceived ideality of the product for customers within a specific market segment. We can apply TRIZ and find ideas, which will improve product functions - we can find ideas, which help customers meeting their objectives better - and we can develop ideas, which focus on value creation for this customer. The three columns (Figure 4) will be used as 'voice of the customer' in the following QFD-examples - just to make things simple.
The A1-matrix can look like Table 4 (4).
Table 4: QFD focused on Product Functions
A1 | Printer speed | Ozone level | Noise-level | Other Characteristics | Customer's Importance Rating |
Fast, high quality printout |
5 | ||||
Low ozone emission |
3 | ||||
Lower noise level | 4 | ||||
Weighted importance: |
58 | 27 | 36 | 4 |
This matrix reflects the customer demand (rows), relevant measurement characteristics (columns), global interaction between these (symbols) and relevance to the customer. The weighted importance (bottom) can help the development team to spend their development resources in a way, which will guarantee meeting relevant customer objectives.
Table 5: Directions to improve the Printer Function
1. Find an alternative way to obtain [the] (faster, high quality printing /satisfier/) that provides or enhances [the] (reviewing source code). |
2. Find a way to eliminate, reduce, or prevent [the] (ozone emission /dissatisfier/). |
3. Find an alternative way to obtain [the] (low noise printing /delighter/) that provides or enhances [the] (accomplishing more complex projects). |
Table 5 lists directions to improve the printer functions for this market segment - they are mainly design related. Examples are: 1) a pretty printer, which prints unstructured source code in a nice format; or 2) a green printer, which no longer emits ozone or 3) a whispering printer.
The middle part of Figure 4 is now taken as customer requirements (Table 6) - we have to adapt the quality characteristics. Some quality characteristics from the previous QFD might still be useful, like "printer speed". The directions of improvement now give non-product related proposals, which may be hard to perform by the printer itself (Table 7)
Table 6: QFD focused on Customer's Objectives
A1 | Printer speed | Average project time | Review frequency | Other Characteristics | Customer's Importance Rating |
Reviewing source code | 5 | ||||
Start more complex projects | 3 | ||||
Weighted importance: |
45 | 27 | 45 |
Table 7: Directions to improve Accomplishment of Customer's Objectives
1. Find an alternative way to obtain [the] (reviewing source code) that offers the following: provides or enhances [the] (functional program with fewer bugs), does not require [the] (faster, high quality printing /satisfier/), is not influenced by [the] (ozone emission /dissatisfier/). |
2. Find an alternative way to obtain [the] (accomplishing more complex projects) that offers the following: provides or enhances [the] (programs with high functionality), does not require [the] (low noise printing /delighter/), is not influenced by [the] (ozone emission /dissatisfier/). |
As manufacturer, consider entering the completely different market of software engineering! Or supply your printer-model with additional benefits. For example, provide a booklet with the printer, 'how to avoid bugs'. The booklet might help the customer reducing the need to print out frequently for checking purposes.
Finally, the supplier might even decide to take a completely different route and support the customer to achieve its ultimate values (Table 8).
Table 8: QFD focused on Customer's Values
A1 | Printer speed | Review frequency | Comfort of Compiler/ Debugger | Other Characteristics | Customer's Importance Rating |
Functional program with fewer bugs | 5 | ||||
Create program with high functionality | 3 | ||||
Weighted importance: |
14 | 42 | 72 |
Table 9: Directions to improve Value Creation for the Customer
1. Consider transitioning to the next generation of the system that will provide [the] (functional program with less bugs) in a more effective way and/or will be free of existing problems. |
2. Find an alternative way to obtain [the] (functional program with fewer bugs) that does not require [the] (reviewing source code). |
3. Consider transitioning to the next generation of the system that will provide [the] (programs with high functionality) in a more effective way and/or will be free of existing problems. |
4. Find an alternative way to obtain [the] (programs with high functionality) that does not require [the] (accomplishing more complex projects). |
Why not providing software engineering services and a printer?
In reality the author made a decision, which matches the last QFD: he switched from a Windows-based environment for software development to a Linux based one. The Windows environment requested using an inconvenient, complex and hard-to-understand library. Linux offered Perl and Perl/Tk. His development rate of complex programs increased considerably, demanding for less individual printouts.
The proposed TRIZ-model can help understanding the customer. It structures the available information into different levels of requirements. This allows doing QFD's on different levels, opening new opportunities for manufacturers. It creates customer-focused ideas, which can be used in many ways - e.g. for concept selection.
The author likes to thank Svetlana Visnepolschi, Ideation International, for her comments on the previous version of this paper (5).