Sample frameDay 245·Product · Translation

Good morning. A customer says it should feel solid. Somebody now has to turn that into a number.

Product.

The House of Quality

5 min read·Apply by lunch

The question

Your customers keep asking for the same vague thing — that it feel faster, or simpler, or more solid — and delivering it means moving one measurable characteristic inside the product: can you say which one, and in what unit? Most teams can't. So the unit gets chosen by whoever writes the code, at the moment they write it.

The idea

In the Harvard Business Review of May-June 1988, John Hauser of MIT's Sloan School and Don Clausing, an engineering professor at the same institution, published 'The House of Quality' — the article that carried quality function deployment out of Japanese manufacturing and into Western product work. The method was Yoji Akao's, and its first large-scale use came in 1972 at Mitsubishi Heavy Industries' Kobe shipyard, after which Toyota and its suppliers developed it further; by the time the article ran, Ford alone had more than fifty applications of it. The house is a grid. Down the left side go customer attributes in the customer's own words. Across the top go the engineering characteristics you can measure and move. The body records which characteristic affects which attribute, and how strongly. Their worked example is a car door: 'easy to close' carried an importance of 7, and translated into the engineering characteristic 'energy needed to close the door', with a target of reducing that to 7.5 foot-pounds. Then comes the roof — the triangular matrix sitting on top — which records which engineering characteristics fight each other. Lower the energy needed to close the door and you weaken the seal and let in road noise. You can't have both, and the roof is what makes you choose rather than discover it late.

The framework

01
Whose words go down the left side?
Customer attributes go into the house exactly as they were said — 'doesn't leak in the rain', 'easy to close', 'I can find things again'. The temptation is to tidy them into internal categories on the way in, and every tidy costs you information, because the specific complaint is the part that tells an engineer where to look. Group them into a shallow hierarchy by all means, but keep the original sentence attached, and rate each one for importance; a house with no weighting turns every attribute into an equal claim on the budget. Importance here is a claim on engineering effort rather than a satisfaction curve — a must-be attribute rates high and still caps out at adequate however hard you push it — so run Day 217's Kano sort across the list before you let these weights drive the grid.
02
An engineering characteristic is anything you could put a unit against this week.
Across the top go the things that can be measured and changed: force in foot-pounds, seconds to first response, decibels, steps to complete, grams. The discipline is the unit. 'Better onboarding' is not an engineering characteristic; 'minutes from signup to first successful action' is, because you can measure it today, argue about the target, and know afterwards whether you moved it. Anything you can't put a unit against doesn't belong on the top row — and the customer attributes left with no measurable characteristic beneath them are the real finding of the exercise.
03
The roof, and the two things you can't have at once.
The roof is the part most people skip and the only part that surprises anyone. For each pair of engineering characteristics, mark whether improving one helps or hurts the other. In the original example, reducing the energy needed to close a car door runs directly against seal resistance and road-noise reduction — you can't maximise all three, and pretending otherwise just pushes the conflict downstream to whichever engineer finds it last. Every negative mark on the roof is a trade-off somebody will make; the roof decides whether it's made by the team or by accident.
04
Benchmark the attributes, then set targets on the characteristics.
Two more rows earn their keep. On the right, score yourself and your competitors on each customer attribute as customers perceive it — that shows you where you're already good enough and where a gain would actually be noticed. Along the bottom, record the objective measure of where you and your rivals sit on each engineering characteristic, then set a target value. The finished sheet reads as a specification rather than a wish: this attribute matters this much, moves through this measure, sits here today, and needs to reach there by release.

Members only · 4 principles + template + AI mentor

364 more frameworks are waiting.

The House of Quality is Day 1 of 365. One framework every morning for a year — across strategy, sales, negotiation, leadership, and more.

$1 /day

Billed $365/yr · cancel any time

365 frameworks

12 topics · templates · AI mentor

Reserve your place — $365/yr →

One payment a year · the whole library · no auto-renewal.

More Product frameworks

All 22 product frameworks →