Quality Function Deployment (QFD) and the House of Quality: Translating Customer Needs into Engineering Specifications

The voice of the customer

The voice of the customer (VOC) is the structured process of capturing customer needs, expectations, and priorities and translating them into meaningful input for the product development process. VOC research typically uses multiple collection methods: in-depth interviews, focus groups, customer surveys, observation of customers using the product or a competitive product, analysis of warranty claims and customer complaints, and review of service records. The goal is to surface both stated needs (what customers say they want) and unstated needs (what customers assume they will get without saying so) and to capture them in the customer’s own language rather than in technical terms.

The raw output of VOC research is a list of customer requirements expressed in customer language. These requirements are then organized into a hierarchical structure using affinity diagrams or tree diagrams that group related requirements and distinguish between primary requirements (broad categories), secondary requirements (more specific needs within a category), and tertiary requirements (detailed, specific requirements that can be measured). This hierarchical organization is the input to the rows of the House of Quality matrix.

One of the most important outputs of the VOC phase is the importance rating assigned to each customer requirement. Customers assign relative importance weights to their requirements through survey instruments (typically Likert scales or constant sum methods), and these weights drive the prioritization of engineering effort throughout the QFD process. Requirements that customers weight as most important receive disproportionate attention in the engineering translation process, ensuring that design trade-offs consistently favor what matters most to the customer.

The House of Quality structure

The House of Quality is a matrix with a distinctive structure that earns its name from the triangular correlation matrix at the top that resembles a roof. The full structure has six interconnected sections:

Left wall (customer requirements): the rows of the central matrix list the customer requirements captured in the VOC phase, organized hierarchically. Each row also shows the importance rating for that requirement.

Right wall (competitive evaluation): for each customer requirement, the competitive evaluation compares the organization’s current product against two or three key competitors on a five-point scale. This competitive benchmarking shows where the product is strong relative to competition (no need to invest more here) and where it is weak (potential competitive advantage from improvement). The competitive evaluation also captures the organization’s target performance for the next product generation.

Ceiling (engineering characteristics): the columns of the central matrix list the engineering characteristics (also called technical requirements or design parameters) that the engineering team controls and can measure. Engineering characteristics must be expressed in measurable, technical terms: not “easy to open” but “door opening force (N)” or “hinge breakaway torque (N·m)”. The engineering characteristics are the team’s language for what they will design to satisfy the customer requirements.

Roof (correlation matrix): the triangular matrix at the top of the HOQ shows the relationships between pairs of engineering characteristics. Positive correlations (two engineering characteristics that support each other) are marked with a plus sign or filled circle; negative correlations (trade-offs where improving one characteristic worsens another) are marked with a minus sign or open circle. Strong negative correlations highlight the most difficult engineering trade-offs and often point to where innovation is needed to satisfy multiple customer requirements simultaneously.

Central matrix (relationship matrix): the body of the HOQ is a matrix where each cell shows the relationship between a customer requirement (row) and an engineering characteristic (column). The relationship is rated as strong (9 points), moderate (3 points), weak (1 point), or none (empty). This relationship matrix is the core analytical tool of QFD: it shows which engineering characteristics most influence each customer requirement and which customer requirements are addressed (or not addressed) by each engineering characteristic.

Floor (technical benchmarking and targets): below the relationship matrix, the floor section shows: the technical benchmarking of the organization’s current product and key competitors on each engineering characteristic; the direction of improvement (larger is better, smaller is better, or target is best); the technical importance ranking calculated by multiplying the customer importance weights by the relationship values in each column; and the engineering targets set for the next product generation.

Calculating technical importance

The technical importance of each engineering characteristic is calculated by multiplying each relationship value in its column by the customer importance weight for the corresponding row, then summing these products. This calculation propagates customer priorities through the relationship matrix to produce a ranked list of engineering characteristics by their impact on customer satisfaction. Engineering characteristics with high technical importance scores have strong relationships with many high-priority customer requirements: these are the design parameters that deserve the most engineering attention and testing effort. Engineering characteristics with low technical importance scores have weak or no relationships with customer requirements: these parameters may be candidates for simplification or standardization to reduce cost without affecting customer satisfaction.

The four-phase QFD cascade

The full QFD methodology extends beyond the single House of Quality into a cascade of four linked matrices that progressively translate customer requirements into actionable manufacturing specifications:

Phase 1 (Product Planning Matrix): the House of Quality described above, translating customer requirements into engineering characteristics with target values.

Phase 2 (Product Design Matrix): the engineering characteristics from Phase 1 become the rows (inputs), and the design characteristics (component-level design parameters, such as material properties, geometry, and tolerances) become the columns (outputs). This matrix translates system-level engineering requirements into component-level design specifications.

Phase 3 (Process Planning Matrix): the design characteristics from Phase 2 become the rows, and the key process parameters become the columns. This matrix translates component-level design specifications into the manufacturing process parameters (machine settings, process temperatures, feed rates, curing times) that must be controlled to produce the component to specification.

Phase 4 (Production Planning Matrix): the key process parameters from Phase 3 become the rows, and the quality control actions (inspection methods, control chart parameters, process capability requirements) become the columns. This matrix translates process requirements into production control specifications: which parameters to monitor, how to measure them, and at what frequency.

Each matrix in the cascade uses the same structural logic as the House of Quality, but the content of the rows and columns changes to reflect progressively more detailed and operational requirements. The cascade creates a traceable chain from customer voice to production control action, enabling engineers to justify every control plan entry by tracing it back to a customer requirement.

QFD in APQP and DFSS

In automotive APQP (Advanced Product Quality Planning), QFD is used in Phase 1 (Plan and Define Program) to translate customer and regulatory requirements into product design requirements that drive the design FMEA, design verification plan, and product design targets. The HOQ provides the systematic structure that ensures all significant customer requirements have corresponding engineering characteristics with measurable targets before the design is committed.

In Design for Six Sigma (DFSS), QFD is used in the Define and Measure phases to identify the critical-to-quality (CTQ) characteristics that must be achieved for the product to satisfy customers. The HOQ prioritizes the CTQ characteristics so that the subsequent Analyze, Design, and Verify phases focus on the design parameters that most affect the highest-priority customer requirements. In DFSS, the HOQ also provides the basis for the transfer function analysis that connects process inputs to CTQ outputs.

Common errors in QFD implementation

• Confusing solutions with customer requirements. Customer requirements must be expressed in the customer’s terms (what they want and why), not in engineering terms (how to deliver it). “Provide a USB-C charging port” is an engineering solution, not a customer requirement. The underlying customer requirement is something like “ability to charge with my current cables.” Expressing customer requirements in solution terms locks out alternative solutions and defeats the purpose of the HOQ analysis.
• Filling every cell in the relationship matrix. A relationship matrix with no empty cells signals that every engineering characteristic affects every customer requirement, which is almost never true. Overpopulated matrices create the false impression that all engineering parameters are equally important and dilute the analytical value of the technical importance calculation. Cells should be populated only when there is a genuine, meaningful relationship.
• Treating QFD as a one-time exercise rather than a living document. Customer requirements change as markets evolve, competitors introduce new products, and regulations change. A House of Quality that is completed at the start of a program and never updated becomes a historical artifact rather than an active planning tool. QFD documents should be revisited at major design reviews and updated when significant customer feedback or competitive intelligence changes the requirements landscape.
• Stopping at Phase 1 and never completing the cascade. The Phase 1 HOQ is valuable, but the real benefit of QFD is the complete cascade from customer requirements to production control specifications. Organizations that complete only Phase 1 miss the systematic translation of design targets into process parameters and quality control actions that makes QFD a comprehensive quality planning methodology.

Frequently asked questions

What is QFD and why is it used in product development?

Quality function deployment (QFD) is a structured methodology for translating customer needs and expectations into specific technical requirements at each stage of product development. It is used in product development because the most common cause of product failure in the market is not poor engineering execution but rather poor translation of customer needs into engineering requirements: products are built to specification but the specification does not capture what customers actually need. QFD provides a systematic process for capturing the voice of the customer, expressing those needs in measurable technical terms, benchmarking the current design against competition, identifying the engineering characteristics that most affect customer satisfaction, and setting engineering targets that will deliver a competitive product. By making the connections between customer requirements and engineering decisions explicit and visible, QFD reduces the risk of building a product that is technically correct but commercially unsuccessful. QFD also improves cross-functional communication: the House of Quality is a shared document that marketing, engineering, manufacturing, and quality can all read and contribute to, reducing the silo-based misunderstandings that often degrade product development programs.

What is the difference between customer requirements and engineering characteristics in the HOQ?

Customer requirements (the rows of the House of Quality) express what customers want in the customer’s own language, without specifying how the product should deliver it. They describe outcomes, benefits, and experiences: “opens easily,” “does not rattle,” “lasts a long time without recharging.” Engineering characteristics (the columns of the House of Quality) express how the product delivers the customer requirements in the engineer’s language, as measurable technical parameters the design team controls: “door opening force (N),” “acoustic panel resonant frequency (Hz),” “battery capacity (Wh).” The distinction matters for two reasons. First, expressing customer requirements in engineering terms prematurely commits to a solution and closes off alternatives. Second, engineering characteristics must be measurable so that targets can be set, designs can be tested against those targets, and the correlation between engineering performance and customer satisfaction can be verified. The relationship matrix in the center of the HOQ quantifies the connections between these two languages, enabling engineers to identify which technical parameters they must control to deliver the customer requirements that matter most.

How is the technical importance of each engineering characteristic calculated?

The technical importance of an engineering characteristic is calculated by multiplying the relationship value in each cell of its column (strong = 9, moderate = 3, weak = 1, none = 0) by the customer importance weight for the corresponding row, then summing all these products. The result is a weighted sum that reflects how strongly each engineering characteristic affects the engineering characteristics that customers care most about. For example, if “battery capacity” has a strong relationship (9) with “long battery life” (customer importance weight 5) and a moderate relationship (3) with “lightweight” (customer importance weight 3), its technical importance contribution from these two requirements is (9 × 5) + (3 × 3) = 45 + 9 = 54. After calculating technical importance for all engineering characteristics, the values are ranked or displayed as a histogram so the team can see at a glance which engineering parameters drive the most customer value. High technical importance scores direct engineering investment toward the characteristics that will produce the greatest improvement in customer satisfaction for a given level of technical effort.

What does the roof of the House of Quality show and why does it matter?

The roof of the House of Quality is a triangular correlation matrix that shows the relationships between pairs of engineering characteristics. Each cell in the triangle represents the interaction between two engineering characteristics and is filled with a symbol indicating whether the interaction is strongly positive (both improve together), weakly positive, no relationship, weakly negative (improving one slightly degrades the other), or strongly negative (a direct trade-off where improving one significantly degrades the other). The roof matters because it exposes the engineering trade-offs that must be managed during design. Strong negative correlations indicate that satisfying two different customer requirements through their respective engineering characteristics is in tension: making the door lighter (to satisfy “lightweight”) conflicts with making it more rigid (to satisfy “does not rattle”). These conflicts must be resolved through design innovation, component-level trade-off decisions, or by accepting that one customer requirement will be partially sacrificed to satisfy a higher-priority requirement. Without the roof analysis, these trade-offs often remain invisible until late in the design process, when they are much more expensive to address. Early identification of strong negative correlations drives innovation effort toward the areas where creative solutions deliver the most customer value.

How does QFD relate to FMEA and the control plan?

QFD, FMEA, and the control plan are complementary tools that address different questions in the product development process. QFD answers “what does the customer need and what engineering characteristics must we control to deliver it?” FMEA answers “what can go wrong with this design or this process, and how do we prevent or detect it?” The control plan answers “what process parameters and product characteristics must we monitor during production, and how?” The connection between these tools flows in a logical direction. The engineering characteristics and targets from QFD Phase 1 define the product characteristics that matter most to customers. These characteristics become the focus of the design FMEA, which identifies failure modes that could prevent the product from meeting those characteristics. The key process variables identified in QFD Phase 3 become the subject of the process FMEA, which identifies failure modes in the manufacturing process. The detection and prevention actions from the FMEA feed into the control plan, which defines the inspection methods, frequencies, and reaction plans for the characteristics and process parameters identified as significant. In practice, an organization that completes the full QFD cascade before drafting the FMEA produces a more comprehensive and better-prioritized FMEA than one that drafts the FMEA without QFD, because QFD ensures that the FMEA addresses the engineering characteristics most connected to customer requirements.

What is the Kano model and how does it complement QFD?

The Kano model, developed by Noriaki Kano in the 1980s, classifies customer requirements into three categories based on how they affect customer satisfaction. Basic requirements (must-be quality) are features customers take for granted: their presence does not increase satisfaction, but their absence causes strong dissatisfaction (a car that starts every time). Performance requirements (one-dimensional quality) are features where more is better: customer satisfaction increases linearly with performance level (fuel economy). Excitement requirements (attractive quality) are unexpected features that delight customers when present but do not cause dissatisfaction when absent (a feature customers did not know to ask for). The Kano model complements QFD by providing a richer picture of how customer importance weights should be interpreted. A basic requirement may receive a relatively low importance weight in a survey (customers assume it will be met) but actually represents a hygiene factor that must be met before any performance improvement matters. An excitement requirement may receive a low weight because customers cannot articulate its value before experiencing it. Understanding the Kano category of each requirement allows the QFD team to set more nuanced engineering targets: basic requirements need threshold performance, performance requirements need competitive parity or superiority, and excitement requirements offer opportunities for differentiation.