Quality is a complex system

Quality is a complex system. Quality is all about roles, relations, context, budget and the question who owns problem.

If you read this blog post, most probably your personal understanding of “quality” is anyway more advanced than mine. I still try to catch up with about 33 years of stubborn ignorance. So feel free to stop reading at any point, if you come to the conclusion: “Well, yes, duh…”

There is something that bugs me now for a few years. It’s the way people treat the term “Quality”! And I used to be one of them.
I have tried many times over the past 5 years to come up with a good and comprehensible explanation of the issues that I have with it. Attempts 1 – 3 ended up in draft mode. This is attempt number 4 and 5! It was supposed to be simple and short and spot on. Guess what: it’s not!

Throughout my life and my career I have met people and seen many examples where quality is treated as a simple thing. It only happens since a few years now, that I’m kind of allergic to sentences like:
“This t-shirt is of good quality!” Is it really? Why? What exactly do you mean? For whom? You want to buy it?

In a professional context my goal for *cough* years is to help improve “quality”. Just do this and it will become better, they said!
Since several years now, I treat these statements with more respect, and many times I begged to differ, and here is why!

When you look at the five word long definitions for “Quality” from some famous people, you could think it is actually a simple thing.
“Quality is conformance to requirements” (Philip Crosby), “Quality is fitness for use” (Joseph Juran) or “Quality is value to someone” (Jerry Weinberg). Sounds easy, doesn’t it? The problem is, that behind those easy statements hides a complex system, that we too often neglect for the sake of simplicity and the urge to be under control.

TL;DR: Quality is all about roles, relations, context, motivation, budget and the question who actually owns problem.

I want to explain now why I think Quality is a complex system.

Quality describes the result of an evaluation of a solution.
When you think about it, every thing that mankind produces or uses is trying to solve some kind of problem. This solution might be something real or virtual, it might be a thing or a process to create a thing. Every thing tries to solve a problem. Quality is used as a way to describe how good this solution is.

What is the definition quality?

According to the Oxford Dictionary, Quality is:

  1. The standard of something as measured against other things of a similar kind; the degree of excellence of something.
  2. A distinctive attribute or characteristic possessed by someone or something.

That is rather vague and imprecise, as it’s not naming any property it refers to explicitly. The term “quality” is actually not able to state anything by itself. You either need to set it into relation to something comparable (#1) and you need to be more specific about the relevant characteristics of the something that you want to refer to (#2).
The term “quality” itself means nothing and everything at the same time. Quality refers to individual characteristics of something, but also means the sum of them all.

And guess what, despite the strong use of the term quality, there’s nearly always a shallow agreement on what it actually means. We use a simple word to refer to a subjective evaluation of a complex system. That has to go wrong…

What means complex?

I’d like to borrow the description of the domains of the Cynefin sense-making framework.

  • Unknown – is the state when you don’t know yet in what domain you are.
  • Obvious/Simple (Sense – Categorize – Respond) – When you do A, B will be the obvious result. A simple and clear set of rules describes what happens in the system.
  • Complicated (Sense – Analyze – Respond) – When you do A and B, C will follow, as long as D is fulfilled. You are still able to formulate rules that describe what will happen in a system. They are just more – well – complicated.
  • Complex (Probe – Sense – Respond) – You do A and B to change C, now also D and E happened. You are able in hindsight to understand why D and E happened. The next time you do A and B, C might change again, but if D and E will also happen or not, you probably still don’t know!
  • Chaos (Act – Sense – Respond) – D happens, you don’t know why, you try A, B, and C. Eventually something about D changes. If you try it again it doesn’t. Now E happens. Where did that come from?! Aaaaah, PANIC MODE! Stay calm, try to understand where you are, get back into the complex domain. Get some control back.

An important aspect about Cynefin is, that it doesn’t try to describe one state for the whole thing. It tries to help you understand which aspect of some thing should be seen in which context, which in turn helps to understand how to treat it. You don’t treat something complex the same way as something simple. You can try, but you have been warned, there might be consequences you haven’t thought of! (Careful, Chaos waits behind the cliff of complacency, if you treat things as too simple!)

The different aspects of quality should be treated with respect to their domain.

What is a system?

According to Oxford Dictionary
A set of things working together as parts of a mechanism or an interconnecting network; a complex whole.

According to Wikipedia:
A system is a group of interacting or interrelated elements that act according to a set of rules to form a unified whole. A system, surrounded and influenced by its environment, is described by its boundaries, structure and purpose and expressed in its functioning.

In my opinion Quality fulfills all the aspects of a system.
A statement about quality is not “just” one single thing, even if it is often treated like that, because we think that we “just” evaluate the quality of a single thing. It’s actually taking into account a list of properties, people, relations, motivations, experiences, and context. People who are involved in the system, all have a different relation to the thing itself and also to all parts that are involved. There are feedback loops between properties that influence each other, changes to the system following different motivations. Each influence changes the system. Many moving parts, many influences from the in- and outside. Relations and motivations you didn’t think about before! Boundaries might change, sometimes unexpectedly.
How big is the system you are looking at? Is it the relation with one single person, or even a long list of people. Are the people all following the same or rather a similar motivation, or do we have properties or key players with competing motivations? Can you look at people individually, or do they have a relation in regards to the thing. What about the process involved to create the thing. Or is the thing a process?
If I evaluate the quality of a thing once, will it stay like that? No, because time elapses and contexts change.
Important: Quality is not even a closed system!

Quality is a relationship between a person and a product/process that varies over time. Yet we usually don’t treat the word as such.

Relations and subjectivity make the difference

What is your relation to the thing and its system, which’s quality you want to describe?
There are objective criteria to evaluate quality, yet we don’t know about the personal order of preference of everyone relevant involved. We don’t know how much weight each aspect of the thing adds to the individual’s evaluation of quality. Do you see it the same? Probably not! Even if you refer to one quality characteristic specifically, your evaluation might still differ.
That’s not the only subjective part, when looking at quality. There can be many relations between quality criteria of a thing and the person using the thing. These relations can be obvious or obfuscated. (My daughter uses an old pair of compasses for math. She prefers that one over the newly bought, because it’s her mother’s. To anyone else, it would just come down comparing two pairs of compasses, that special relation is missing in their evaluation.)
Relations are the key aspect in my opinion that make the system of quality complex. You cannot understand ALL of them, so you can never predict what will actually happen in this large net of intertwinediness .

Who owns the problem?!

Like aforementioned, every thing used and produced by mankind tries to solve a problem. That even includes the process to produce things. Some problems are more obvious than other. (A fork is usually kind of obvious. What about a piece of decoration for the mantelpiece?)
When we want to evaluate quality, we need to know, what part of the solution are we looking at, who owns the problem, what is their relation to problem and solution, and what is their motivation to solve that problem?

When the potential consumer base is large, you can’t create the perfect solution for each individually. To reach the maximum audience, you need to find a way to disappoint as few as possible. The system has to reach a certain level of balance.

What is your role?

Do you own the problem and want to solve it, so you evaluate potential existing solutions or proposals?
Are you aware of a problem and want to solve it? Whom do you solve it for? For yourself, for someone else, someone you know and have access to for feedback, for yourself and others, only for others? In case more than one party is involved, is the problem really the same? Is everybody feeling the same about the relevance of the problem. Does everybody think that the proposed solution is a good solution? Is your solution competing with others?

Motivation and budget

Someone wants to solve a problem and evaluates one or more solutions and comes to a decision about the best candidate available.
But there are other aspects attached to the problem as well: motivation and budget! Do I want to solve the problem? Can I and do I want to afford this thing to solve my problem. This is not only about money, also about time, space, and everything else that is scarce (which is basically everything!).
Is solving the problem worth to acquire this solution?! Is my motivation to solve the problem so high, that I take the next best solution? Do I want to spend the amount and effort?
If the person who owns the problem decides: No! Then your solution is not a solution, but only a proposal.
If your interest is to “sell” your solution, you need to keep that in mind. And it’s not only about money. The “price tag” might also mean prestige, or positive feedback that rewards you.

There is also the situation that someone has to use a solution that someone else decided to be valuable for solving a problem. Think about the software system you have to use at work. Think about the pullover your mother bought you when you were eight. How does the evaluation of quality change for you and them. What role does the budget play?

From a kind of business perspective, the only quality evaluation of the thing that counts, is the one from the acquirer. You need to convince potential consumers that your solution is worth it. (That’s the reason for shiny marketing brochures to convince management, not the actual users!)
I can produce the best solution in the world. If I don’t find a problem owner who wants to use it, it’s probably worthless.
I can think highly of a certain product/process and praise it. As long as I don’t have a problem that it fits to or I just buy it because, you know, reasons…, what is my role?
I may be an influencer for someone who has to make a decision about acquiring a solution, and suddenly I’m part of the system and my opinion counts.


Working effectively in the “Quality” domain means to understand and never underestimate the system’s complexity. You should know about as many relevant aspects, and be able to predict what could happen when you change a thing. For example, adding security to anything usually results in reduced performance and usability. Are your consumers really preferring additional security over speed and ease of use?

You need to understand the relations between quality aspects as well as their relation with stakeholders, as well as many other relations, like for example between a potential user and the producer. Is the brand image important to a potential consumer of the solution? Is “having” more important than actually solving a problem? Know and understand your audience!

You need to understand the problem and people who have the problem. You need to help understand the impacts on quality in case of certain changes. You need to balance the quality and the resulting price tag with the budget of your potential users. You can put gold plating on a sledge hammer. But don’t prepare for mass production just yet!
The quote “Just because you can, doesn’t mean you should!” comes to mind. Both from an ethical and a business perspective.

Be warned, I might add a second part to this article using a couple of real world examples and explain the difficulties when it comes to evaluating quality.


Introducing Grumplexity Theory

In close relation to Prof. Dave Snowden’s Cynefin framework I found out this week, that the model also fits very good to describe the states of grumpiness.

Being grumpy is a state of mind that I – and also other fellow testers, like my partner in grump, Del Dewar – have internalized. The reason for the grumpiness though is highly context-dependent and should at no time be underrated. The grumplexity model provides you a tool, to understand your own and others’ grumpiness a bit better.

When you enter a situation, you know the person is grumpy, but you have no idea why. You don’t know yet, how to deal with the situation, and if you should at all.

Bildschirmfoto 2018-02-23 um 15.41.06

Sometimes the reason is quite obvious. The train didn’t come, it’s raining and there’s no cover. After the trigger has been removed, solved or vanished into thin air, the chances are good, that a piece of chocolate might help.
I call this the grumpvious mood. The half life is usually not very long, once the trigger and most consequences are gone or solved. The train finally came, you found shelter, your clothes start to dry again. A piece of chocolate could help.

Then we come to the situation where it’s not that simple. The trigger might be a bit in the past, the initial situation is already forgotten, so it seems. But real grumpsters don’t forget.
Let’s take a work example. Someone broke the build process, because they forgot to build it locally before checking in; for like the hundredth time. Then a few days after the last time that person broke the build chain, you see them check in again, without building it locally. You get grumpy, because you know of the apparent risk, and can’t believe that this person still hasn’t learnt from the past hundred times.
Nothing happened – yet – but you are grumpy already. I call this a grumplicated situation. Not directly apparent, but easy to explain. A whole bar of chocolate might help in that situation.

Now it get’s tricky. It’s for example about triggers that don’t seem to be qualified to make you grumpy in the first place or multiple triggers that seem independent and suddenly become dependent. For observers it’s even harder to understand what has happened to trigger that mood.
Let me give you two examples. When I come home tonight, I’ll have to help my wife bake some cake. Nothing bad about that, but actually I have to do the baking, because she broke her hand a few weeks back, and the reason for baking the cake is not for self-consumption, but for a promise my wife gave before her accident. And I’d rather go into my workshop to continue the reconstruction I started weeks ago, before my wife had her accident, and that didn’t make much progress. Yet, I’ll of course help her with the cake and get up early on a Saturday to deliver it. That is a grumplex situation, as multiple triggers come together. And it might leave the colleagues in wonder, why I’m going home grumpy on a Friday afternoon.
The other example for a grumplex situation, and the trigger that started this whole grumplexity idea, was an email I received recently with an actual positive content. The problem was the way that lead to the situation and my involvement in the process, and all the strings attached to this, that instantly made me heavily grumpy and brought up the urge to reply in a very unpolite manner. Grumpsters don’t forget easily. Now imagine the perplexity of the sender of the email, if they would have received the unpolite reply.
You better bring some more chocolate and a nice bottle (or two) of Spezi (my favorite drink), if you want to have a proper conversation with me in that situation.

And last but not least, there are the days where you are just batshit angry, just because. It takes some energy and lots of chocolate and Spezi to calm down and find the energy to analyze the situation to unscramble everything that goes wrong at the same time. That is when a bunch of triggers made an appointment to come up on the same day. Individually it might be able to handle them, but coming at you all together. Holy shit, you better duck.

Update: David Högberg sent me this link that describes how easy it is to reach batshit angry mode. hyperboleandahalf.blogspot.se/2010/05/sneaky…

And then there is special situation that I call the cliff of fake happiness. You try to keep your grumpiness to a minimum, and it seems nearly like you are smiling. But you should better wear a helmet near a grumpster, when they smile. Never feel safe in such a situation, the next moment, you don’t even know what hit you; an empty bottle of Spezi, a wrong meowing of the cat, push message on the smart phone. Boom!!!

This article is supposed to be fun to read in the first place, but it also could help you a bit with understanding the different domains of the original Cynefin model, and it helps you to understand why a grumpster is grumpy and that often a piece of chocolate is not enough and might not last long.

I want to thank Zeger Van Hese for adding the term grumplicated and giving me the spark to this.