“No-one knows the cost of defective product. Don’t tell me you do. You know the cost of replacing it, but not the cost of dissatisfied customer.” -W.E. Deming

Lean Management and Value

Deming stands as a central figure influencing Lean principles. His statement about the significance of customer satisfaction encapsulates a crucial idea: by doing things right the first time, a company can align its operations in such a way that forces like recommendations can be harnessed for its benefit. While discussions on net promoter score (NPS) and customer satisfaction are common topics in Zeffi's blog, this time we focus on quality as a production factor and the drivers that influence it.

As mentioned in previous parts of the Lean Blog Series, Lean management aims to achieve quality that meets customer expectations while simultaneously keeping costs in check, with one key method being the avoidance of waste. There's often a shortcut taken, stating that Lean pursues "high quality" – however, this notion misses the mark. Producing quality higher than customer expectations would be value-destroying, thus wasteful in essence.

Quality Assurance with Quality Function Deployment

Quality Function Deployment (QFD) is a method that assists you, as a Lean practitioner, in focusing on what the customer expects. The essential question is what conditions must be met to fulfill the actual needs of a typical customer. This moves us from the realm of product conceptualization, which we are familiar with as product developers, closer to a computational world – objectively identifying which aspects are meaningful value sources to the customer, i.e., what components contribute to the perceived quality of a product or service.

QFD is also referred to by another term, the House of Quality. The reason for this designation becomes clear from the shape of the diagram. While there is no standardized definition for the term "House of Quality," its application can be tailored to specific organizations. In this blog, we address the general components of QFD and implement a digital tool for quality assurance.

Generally, QFD comprises the following factors:
1.    Voice of the Customer (VOC) regarding the importance of different product attributes (weights, e.g., 1-5)
2.    Controlled factors (measurable variables within your company's control)
3.    Relationship between the above (weak, moderate, strong; e.g., 0-10)
4.    Estimated performance of competitors regarding product attributes
5.    Relationships between factors (very negative, negative, positive, very positive)

Figure 1: Example of a Quality function delpoyment (QFD) graph a.k.a. House of quality

Practical Exercise – Listening to the Voice of Customer with QFD

In this functional part of the blog, we determine the weights of Voice of Customer (VOC) quality factors ("what the customer values") and identify which factors have the greatest impact with a multiple-choice question ("what can be influenced"). We use a one-dimensional graphical question (a slider question) to determine the weights of drivers, and then delve deeper by exploring the relationship between drivers and quality factors. This process objectively elucidates the customer's perspective on the core of QFD.

Another practical tool related to QFD, implemented here, involves gathering the perspectives of essential stakeholders, typically internal, on the impact of drivers on quality factors from the customer's viewpoint. Unlike the previous example, we use image choice questions here, selecting the most suitable from the controlled factors. However, this could also be done with linear questions, which according to academic research provide a more comfortable interface for respondents. A linear implementation would increase the number of questions, as each quality factor would need to be assessed for each controlled factor, resulting in, for example, 49 items (7*7) to evaluate – a feasible but rather laborious method even with an electronic tool.

Figure 2: The idea of the Value compass survey is to score the process steps as a basis for discussion and to get ideas about automation or work reorganization

One might initially consider asking customers about this section, but most likely this part of the analysis includes factors that cannot be disclosed to a wider audience. When asked to internal stakeholders (or in a subcontractor model, other stakeholders), factors deemed controllable value drivers can be more openly discussed. The practical example in Figure 2 shows the evaluation of individual factors, asking one at a time about the most significant drivers affecting each quality factor. In this example, the number of drivers is limited to three, as the exercise should focus on the essential aspects, allowing respondents to choose less than half of the available items. You can try responding to this exercise here.

Figure 3: The idea of the Value compass survey is to score the process steps as a basis for discussion and to get ideas about automation or work reorganization

Regarding the evaluation of the House of Quality's roof, no investigation into interdependencies is conducted via electronic survey in this instance. However, it should be noted that this, too, could be delegated to a larger group for processing. Nevertheless, internal stakeholders are used as evaluators, not customers. In this context, a competitor assessment is also not conducted, but it would be a relevant aspect in a full-scale QFD exercise.

Next-Generation Diagnostic QFD with AI Assistance

The aforementioned process offers an objective perspective on the quality delivered to customers, compared to the more traditional method in which you, as an expert, make choices about the significance of quality factors on behalf of customers. However, we haven't completely freed the customer to express their most significant factors based on the traditional Lean playbook. Where did we go wrong? The answer: we evaluated aspects or quality factors on behalf of the customer.

Could this be done differently? Yes, it can. Let's consider another exercise – first, we ask customers about their quality perceptions, and then we let generative AI find the root cause behind their rating and summarize the root causes from all respondents. This way, the process becomes truly customer-driven, particularly in the Voice of Customer section.

Figure 14 The idea of the Value compass survey is to score the process steps as a basis for discussion and to get ideas about automation or work reorganization

Generative AI can assist in extracting the right insights from interviews and complement traditional Lean methods. Additionally, it supports objective classification, enabling the identification of the most significant factors expressed by customers regarding quality. We, as researchers or developers, don't dictate what matters most to customers; they themselves provide the input for the AI classification. The example in Figure 5 illustrates the VOC themes derived from interviews conducted with a hypothetical 20 customers for Company XYZ's products.

Figure 5: The idea of the Value compass survey is to score the process steps as a basis for discussion and to get ideas about automation or work reorganization

Viewed in this light, the VOC process becomes even more customer-driven. AI performs classification and typification typically more reliably and objectively than its human counterpart. Of course, it's essential to exercise discretion, but for the tools of Lean management, it's crucial to remain vigilant and consider whether generative AI could genuinely be an alternative to traditional methods or a complement to them. Preliminary experiences are certainly intriguing and suggest that AI-assisted QFD is highly effective.

Conclusion – Producing Quality, from Whose Perspective?

In this blog post, we have focused on Lean and quality, particularly QFD. It's essential to occasionally challenge ourselves and our organizations – when we talk about quality, whose perspective defines it? If not the customer's, then whose? And how do we avoid optimizing product and service development from the perspective of small, silent minorities?

As a SaaS provider, we are constantly faced with this latest question, as are others in the industry. Thus, development should follow the perspective of a significant majority, which can be (thoughtfully) supplemented with customization options and add-ons. And this perspective isn't limited to the software industry but is universal in business – how do we maximize customer satisfaction?

The systematic principles of Lean, such as QFD presented here, help crystallize the most critical factors. The majority opinion can be sought by crowdsourcing opinion formation from a sufficiently large sample of customers, rather than optimizing based solely on received customer feedback. A different approach could lead to a strong bias in development and optimization from the margins.

Certainly, it's possible and recommended to consider different smaller customer segments (systematically) and examine the impact of various background factors on their choices. This is easier with electronic tools and high-quality reporting.

Hopefully, this article helps you in your reflections on quality! Next time, in two weeks, we will discuss the 8th waste of Lean – wasted talent or knowledge.

Regards,

Janne Vainikainen, Zeffi, COO, Lean Six Sigma Black Belt

Blog Series Overview:

• Part 1 - Lean Management in Practice
• Part 2 - Lean Tools for Identifying Value
• Part 3 - Lean Tools for Ensuring Quality