Anyone who has worked in quality management in a manufacturing company has seen a version of the same sequence. The procedures are current, deviations are logged, the external audit passes without major findings, and the company still loses customers over something those customers themselves call quality. The explanation usually offered is poor compliance, and it rarely survives a look at the records. A quality system governs the distance between the specification and what gets delivered. Quality itself is settled earlier, in the reading of what the customer needed and what the customer is willing to pay for. If that reading is wrong, a functioning system does not make the error smaller; what it does is make the error reliable.
Five meanings, not one
The word quality carries at least five different meanings, and all five are professionally valid. David Garvin gathered them in 1984: the transcendent meaning from philosophy, the product-based one from economics, the user-based one from marketing and operations management, the manufacturing-based one and the value-based one. His purpose was not to name the correct one. He showed that each of them captures something true about quality, and that any one of them on its own gives a limited picture.
The manufacturing-based meaning is conformance to specification, and that is the one a quality system is built to govern. The user-based meaning asks whether the delivery met the need the user actually had, and the value-based one concerns the relation between what is delivered and what it costs. When a quality department reports on conformance alone, the report therefore covers one of five meanings, even though its heading promises all of them.
Where quality is settled, before anyone delivers anything
The mechanism was described in 1985 by Parasuraman, Zeithaml and Berry. Their model holds that the quality a customer experiences corresponds to the distance between the customer's expectation and the customer's perception of what was delivered. What is interesting here is not the conclusion itself but the three gaps on the supplier's side that the model accounts for. The first lies between the customer's actual expectation and management's perception of it. The second lies between management's perception and the specification that gets written down. The third lies between the specification and what gets delivered.
A quality system works on the third gap, and that is why it does what it was built to do: it is a well-documented way of making sure that what was written down is also what was carried out. The first two gaps lie outside the system’s reach, and they lie before it. If the expectation has been read wrongly, or the reading has been translated wrongly into a specification, the system can only ensure that the wrong thing is delivered with high reliability and full traceability.
"A quality system cannot discover that the specification is the right one. It can only discover that it was not followed."
What the certificate actually documents
If certification created quality by itself, the effect ought to be readable in company results afterwards, and it is not readable in any straightforward way. Gavin Dick, Iñaki Heras-Saizarbitoria and Martí Casadesús have examined the direction of causation and point to reverse causality as the main mechanism behind the better results of certified companies: companies that already perform well are more likely to seek certification. The certificate is therefore to a considerable degree a consequence of a competence rather than a source of it.
It is worth being precise here. Certification has documented effects, particularly for the company that undertakes it for internal reasons rather than to get onto a supplier list, and the authors do not dispute that certification works. Their finding is that self-selection explains much of what one would otherwise attribute to the certificate itself. For anyone reading a supplier's certificate, the consequence is that the paper documents an implemented and observed system, not the quality of what the supplier actually makes.
Why the standard is not the problem
All of this could be read as an objection to ISO 9001, and that would be a misunderstanding. The standard itself asks for what is being called for here: that the customer's requirements be determined, and that satisfaction be monitored. It asks, in other words, for the first two gaps. The problem arises in the way the requirement is typically met. Satisfaction becomes an annual questionnaire with an average score, the customer's requirements become whatever the order confirmation said, and the two gaps the model calls decisive end up reported rather than investigated.
That is how a quality function comes to be entirely in order by its own measure and blind precisely where the company loses money. The cause is rarely laziness or a lack of competence. Conformance can be measured tomorrow, whereas a need the customer has not put into words requires someone to go out and ask.
What the customer pays for
The value-based meaning of quality is the one least often measured and the one that most often decides whether the customer comes back. It concerns the relation between what the customer gets and what the customer pays. In practice there are properties of a delivery that the customer pays for without anyone measuring them, and properties that are measured and reported without the customer having asked for them or being willing to pay for them. The first cost customers, the second cost hours, and both are invisible in a report that only accounts for conformance.
This is also where quality and strategy run together. A strategy is a choice about which customers the company intends to be the right supplier for, and quality is the question of whether it then is. That judgement cannot be made inside a management system, because the system does not know the customer's alternative. It requires someone to find out what the customer compares with, and what the customer would give up first if the price went up.
Three questions a management team can answer this week
The questions address the two gaps a quality system does not cover, and they can be answered without starting a project.
- •What did the customer buy? Not what we delivered, and not what the order said, but what the customer needed solved — and who has spoken with the customer about it since the quote went out?
- •Where in our specifications is there an assumption about the customer that nobody has tested? Every specification contains them, and they are rarely wrong to begin with. They become wrong because they stay put while the customer moves.
- •What does the customer pay for that we do not measure, and what do we measure that the customer does not pay for? The second half is almost always the easier one to answer, and it is also where the hours are.
None of the three questions can be answered by a quality system, and none of them is an argument for getting rid of one. A system that governs the distance between specification and delivery is the precondition for being able to trust a specification at all. It simply never answers whether the specification was the right one, and that question is the only one the customer ultimately votes on.
✓What matters most
- →Quality carries at least five professionally valid meanings. Conformance to specification is one of them, and the only one a quality system governs.
- →The two gaps that decide perceived quality lie before the delivery: the reading of the customer's expectation and its translation into a specification.
- →If the specification is wrong, a functioning system does not make the error smaller. It makes it reliable.
- →A certificate is to a considerable degree a consequence of competence rather than a source of it, and therefore documents a system rather than a quality.
- →What the customer pays for without anyone measuring it is the most expensive blind spot a company can have.
Sources
- Garvin, D. A. (1984). What Does “Product Quality” Really Mean? Sloan Management Review, 26(1), 25–43. The five meanings of quality.
- Parasuraman, A., Zeithaml, V. A. & Berry, L. L. (1985). A Conceptual Model of Service Quality and Its Implications for Future Research. Journal of Marketing, 49(4), 41–50. The model of the gaps between expectation, perception, specification and delivery.
- Dick, G. P. M., Heras-Saizarbitoria, I. & Casadesús, M. (2008). Shedding Light on Causation Between ISO 9001 and Improved Business Performance. International Journal of Operations & Production Management, 28(7), 687–708. On the direction of causation between certification and results.
Sources are cited in their original language. Garvin’s article and Dick et al. sit behind the publisher’s paywall; the reference to Dick et al. rests on the article’s abstract and the authors’ earlier work on the same question, not on the full text.