Quality Function Deployment (QFD) is a structured planning method that translates the voice of the customer into measurable engineering and technical requirements, so that what a factory actually builds matches what the market actually wants.
Developed in Japanese shipyards and manufacturing in the late 1960s and 1970s, QFD replaces guesswork and internal opinion with a documented chain of logic that runs from vague customer wishes all the way down to specific process settings on the shop floor.
The centerpiece of QFD is a matrix nicknamed the House of Quality because of its roof-shaped correlation section. This article explains the House of Quality, walks through the four-phase QFD flow, works a real numeric scoring example, and shows how QFD connects to downstream tools like the control plan, FMEA, and process capability.
Fabrico plays a specific role here: it is the real-time data foundation that holds and monitors the critical-to-quality characteristics QFD identifies.
The House of Quality is a single matrix that links customer needs to technical responses. It has six recognizable rooms:
Filled in together, these rooms convert a wish list into a ranked, quantified engineering agenda.
QFD is not one matrix. It is a cascade of four linked matrices where the "hows" of one phase become the "whats" of the next, carrying customer intent all the way to production.
By Phase 4 you land squarely on the shop floor, which is exactly where a monitoring platform like Fabrico measures whether those parameters hold in real time.
Suppose an automotive parts maker gathers three customer wants and rates their importance on a 1 to 5 scale: coating does not peel (importance 5), consistent color (importance 3), and fast delivery (importance 2). Three technical hows are on the table: cure temperature, coating thickness, and line speed.
Each want-versus-how cell gets a relationship score of 9 (strong), 3 (medium), 1 (weak), or 0 (none). The weighted contribution of a technical requirement is the sum of (importance x relationship) across all wants.
Cure temperature relates strongly to no-peel (9), medium to color (3), none to delivery (0):
Coating thickness relates strongly to no-peel (9), strongly to color (9), none to delivery (0):
Line speed relates weakly to no-peel (1), weakly to color (1), strongly to delivery (9):
The ranking is clear: coating thickness (72) is the highest-leverage engineering requirement, followed by cure temperature (54), then line speed (26). To normalize, divide each by the total (72 + 54 + 26 = 152): coating thickness carries 47 percent of the technical priority, cure temperature 36 percent, and line speed 17 percent. That tells the design and process teams exactly where to spend engineering effort and inspection budget.
The roof of the house is where QFD earns its keep. In the example above, raising line speed to satisfy fast delivery may reduce coating thickness and cure time, directly hurting the two most important customer wants. The roof marks that as a strong negative correlation.
Surfacing this conflict on paper, before tooling is cut, is far cheaper than discovering it as field failures. These trade-offs feed naturally into a FMEA , where each high-risk interaction is scored for severity, occurrence, and detection.
The high-scoring technical requirements from the House of Quality become your critical-to-quality (CTQ) characteristics: the handful of measurable parameters that most determine whether the customer is satisfied. In the worked example, coating thickness and cure temperature are the CTQs worth tightly controlling. Two downstream tools take over from here:
Before any of that math is trustworthy, your measurement system itself must be reliable, which is why a Gauge R&R study belongs early in the chain. QFD points the whole factory at the right variables; capability studies prove you can deliver them.
Fabrico does not run QFD workshops or generate the House of Quality for you, and it does not perform statistical software analysis or design-of-experiments on your behalf. QFD is a human, cross-functional planning exercise. What Fabrico does is capture and monitor, in real time, the machine and production data behind the CTQs that QFD identifies.
Fabrico measures live OEE and the quality losses inside it, tracks scrap rate so you can see whether your top-priority characteristic is drifting, and, through computer-vision monitoring, watches machines that have no PLC at all.
Its CMMS ties the maintenance side in: when a cure oven or coating head starts wandering off target, work orders, assets, and spare parts are tracked in one place, supporting the kind of proactive maintenance that keeps CTQs stable.
In short, QFD decides what matters; Fabrico gives you the honest, continuous data that proves whether the line is holding it.
Teams stumble on QFD in predictable ways. They pack the matrix with dozens of wants and hows until it becomes unreadable, when a focused house of the vital few is far more useful. They fill in relationship scores by committee vote rather than evidence.
They skip the competitive benchmarking room, losing the market context that makes targets meaningful. And they treat QFD as a one-time document instead of updating it when real production data, often surfaced through a Pareto analysis of actual defects, contradicts an assumed relationship.
QFD works best as a living link between customer intent and measured shop-floor reality.
No. QFD started in product development, but the same House of Quality logic applies to redesigning existing products, planning services, and improving processes. Any situation where you need to translate what people want into measurable technical characteristics is a candidate, including internal process improvements you might otherwise tackle with DMAIC.
QFD is forward-looking and prioritizing: it decides which requirements matter most to the customer and what your targets should be. FMEA is risk-focused: it asks how each of those requirements could fail and how to prevent or detect the failure. They are complementary. QFD identifies the critical characteristics, and FMEA then stress-tests them for failure modes before you commit to production.
Not to build the matrix itself, which many teams complete in a spreadsheet during cross-functional workshops. What you do need is trustworthy production data to validate the relationships and to monitor the critical-to-quality characteristics once you are in production. That is the role real-time monitoring plays: it does not create the QFD, but it tells you whether the priorities QFD set are actually being met on the line.
Ready to see whether your line is holding the critical-to-quality characteristics your QFD identified? Book a Fabrico demo and watch real-time OEE, scrap, and machine data become the honest foundation your quality planning depends on.
Login
Blog
