Quality = Engineered + Experienced - explained with fountain pens.

Three of my fountain pens (Lamy 2000, TWSBI Eco, Sailor Pro Gear)

I'm about to make a huge mistake by publishing this post, as today's topic is: quality.

Is there a concept, a word, that's more divisive in the software testing community than quality? I have to admit, I have considered writing a book about it. That would have been an even bigger mistake than publishing this post. Glad to report that I didn't even try, and this blog will hopefully be the last thing I'm writing about this topic.

Let me propose a thinking model for quality.

Quality = Engineered + Experienced.

Last week, I posted a controversial take on LinkedIn, namely that I do not believe quality engineering is possible. This post is my attempt to explain my stance in more detail. Feel free to disagree, but at least hear me out.

Here's how I understand "quality engineering", with the help of a fountain pen.

Here's the architecture of a fountain pen:

Every part of the fountain pen should be engineered correctly, or the pen as a whole might not work. There are more crucial aspects to it, for example, if the nib is broken that would render the whole pen useless. Whereas, with a slightly cracked cap, you could still write normally.

This is what I think the people who call themselves quality engineers are focussing on. They're trying to make sure that every bit in the architecture of their software functions well enough, and they also look at the whole. They are then confident enough to declare that their software has good quality. All because they looked at all the quality attributes they consider relevant and took measures to "quality-engineer" those to their satisfaction.

Sure, they'll take user satisfaction into consideration, but apparently, that doesn't stop them from being convinced that they can engineer quality.

I could never. To me, this feels wrong. I could never work on software and declare that the quality is good enough. There's just too much we don't know about our own product to ever make that claim.

Software as a whole just sucks too much. The tech industry does NOT have a good enough track record to be able to claim that we can engineer quality.

Experienced

Speaking of feelings, that leads me to the next part of my model. Quality is experienced.

To go back to the fountain pen example, you can craft a pen with perfect technical quality, and it will still be hated by some people. That's because using a fountain pen is a highly subjective experience. What pen you would like to use depends on your mood, your overall preferences and your other experiences with the brand.

I have a Japanese pen (Sailor Pro Gear) that has a scratchy nib. This is a divisive feature in a fountain pen. Some days I like it, other days I don't. Some people don't like scratchy nibs at all, and for them, most Japanese pens would feel like there's something wrong with them.

Some people avoid certain brands completely because they made weird marketing choices (Lamy chose to create Harry Potter branded pens, leading to some people never buying from them again). To those folks, it doesn't matter that Lamy engineers quality pens, to them the brand is anti-quality (because quality can also concern not supporting a company that does business with a woman who fights trans rights).

Dear reader, you are either on my side at this part of the argument, or you are not.

I'm quite sure that some of you will scoff at the notion that quality is experienced and that you can't engineer this bit, thinking: what does this matter if it's beyond our control? But to me this is the more important part of the equation!

I wouldn't go as far as to say that a pen that is technically well engineered but doesn't fit my mood at an exact moment in time is of "bad quality", I'm only saying that it is part of the quality story AND that you cannot engineer this bit. And that it matters.

And that is, in more words than fit a short rage-baity LinkedIn post, my issue with quality engineering.

Testing focus

I'm okay with the fact that the quality experience is largely out of my hands. I don't feel a need to control everything.

As a tester, I focus on finding problems that threaten the value of the product. A lot of the work I do could be classified as "quality engineering", but I just don't choose that branding.

There's nothing wrong with testing as a role, as a profession. Too many of us have simply bought into the notion that testing isn't enough, and that's a shame. Well, I'm going to buck that trend.

Quality nirvana

To close off, here's a Venn diagram I made with this quality model.

What would you call it, if a product manages to be well engineered and pleasantly experienced? Quality nirvana?

Does it ever happen? Have you ever experienced it with software?

quality = engineered + experienced. What if both these things come together?!

Alternatively, here's my proposed realistic situation:

Most software sucks. Even if testers, or quality engineers (or whatever the fuck you want to call yourselves), are involved.

Instead of talking about quality, which I will now stop yapping about, we'd be better off taking a good hard look in the mirror. We have to figure out how we, as a tech industry, failed our users.

Quality isn't engineered, my god, there's barely any quality at all. It's not even a focus for most companies. Oh sure, they say that they care, but actions speak louder. They only care about money, and focus on extracting as much of it as they can in the short term.

There's not much quality engineering happening in a late stage capitalist world, I'm afraid. Just a minority of people who keep trying to do the good work in a world that largely doesn't care any more. So no, if you call yourself a Quality Engineer, I'm not mad at you or something. I just think you are slightly delusional.

Alright, that was this week's post. I'm going to go back to writing with my fountain pens, which surely puts me in quality nirvana. Cheers.