Context eats Process for Breakfast

That was the title of a workshop that I gave at Let’s Tests 2016 in Runö, Sweden. In my opinion I messed up that workshop, as I still think that I was not able to communicate the intended message to the participants. So I want to give it a new try, over 5 years later, this time as a blog post. Because I still think that the message I wanted to teach back then is important and basically necessary to understand certain aspects of software development.

Breakfast?! Why breakfast? Well, that was the theme of this after lunch workshop.

The first task was given to one participant only. They had to describe to me: “How to make good coffee?” Now comes the tricky part. I don’t drink coffee, I don’t like coffee, I often even can’t stand the smell of coffee. But I want to be able to offer visitors a good cup of coffee. So the volunteer described me several steps, how to get some ground coffee, put it in the machine, add some water, bring it to a boil, let it filter through the coffee and et voila, coffee! So they told how to make coffee.
But how do I, as someone who never drank and will probably never drink one, know that I just made good coffee? The volunteer told me again to stick to the process and I will get good coffee. Okay, but how do I know that I don’t produce bad coffee.
You might guess what my problem here was and what I tried to address.

The next task was split into two parts. First, the participants had to individually “draw how to make toast”. We had some excellent process descriptions and most of all some fantastic drawings, of how to make toast. Then as a small group they had to throw their process steps together to get more detailed steps, add steps that were forgotten earlier. This time the task was “to make GOOD toast”! The results were again very detailed and incorporated some excellent improvements. But nobody was able to tell me why THEIR process produces GOOD toast.

You will probably sit there, reading this post, nervously shaking, as you might see since two paragraphs now, where I wanted to point my participants to. They made great task analysis and gave excellent process descriptions. But in these three rounds of exercises nobody gave me any details on how to create something “GOOD”. They all focused on creating “something”. And I had an excellent audience this afternoon, but they refused to “get it”.

A good friend of mine loves coffee, and I like to chat with him about coffee. (“Know your enemy” is one of my mottos!) So basically I know a lot about certain aspects of the coffee making process that can influence the outcome of the taste of the product. Getting the right beans, grinding them to the right coarseness, right water temperature and pressure, preheated cups, and so on. There are so many small adjustment screws in this process that can influence the taste of the outcome.
And the same goes for making toast. What bread do you choose, thickness of the slices, what kind of toaster, temperature, time, degree of roasting. Probably even, what do you serve with the toast? What aspects make it a good toast?

Or to refine it: what quality aspects define “good” in those contexts. The answer is relatively easy: It depends! On the context and the consumer!
In what environment are you preparing your coffee or toast? For yourself, for friends, do you work at a small breakfast place or in a large hotel? Do you know your consumer, do you have influence on certain decisions, and so on.

Everybody was so busy describing the steps, that they totally forgot about the quality criteria that need to go into decisions in several steps to make something “good”.

The part that had to be removed due to not enough time, was to describe one process from their work context. My favorite example would have been processes, like the bug reporting process, because most places have one of these.

What does a bug reporting process describe. Basically a lot of things, like the tool, the workflow, permissions, which priority to select, which information to provide, and so on. The only aspect I have seen so far in the about a dozen different bug reporting processes of my life, that positively influenced the quality of the bug reports are certain rules about what information to provide in a ticket. You can google for that and you will find some great examples.
But what makes a good bug report? When I know the process how to create the ticket and whom to assign it to? When I fill out all the necessary information? Well, that makes it a good ticket for the bug report management. That is only one aspect of the whole process. But is the individual result really a good bug ticket?

How much pre-analysis have been made? Did you add log files, minimum reproduction steps, test data and so on. Did you add some bug advocacy, e.g. business reasons why this bug needs to be fixed with the given priority. Who is your audience and will read your bug report and has to understand it? Of course there are contexts, where you already spoke with the dev and you agreed to enter a bug ticket to document the fix. These tickets will probably hold less information, because they are “good enough” in that context. In other situations you know that the bug won’t be fixed now, but only later. And you don’t know by whom. How do you decide then, what is “good enough”?

That is the message that I wanted to communicate 5 years ago, so here you get it now in all explicitness.
Your context and your consumer influence the “goodness” of your product. You can describe the best step-by-step instructions, but if you are not aware of all the small factors that change the quality of your product, and if you are not aware of who your customer is, and what they value, then your process will only ever describe how to get a product, but never how to get a good product.

As a big fan of the AB Testing podcast and the Modern Testing approach that evolved from the discussions between Alan Page and Brent Jensen, I’m very happy that Modern Testing principle #5 describes exactly that message:

#5 We believe that the customer is the only one capable to judge and evaluate the quality of our product.

You can cook a good coffee for yourself, but if your guest doesn’t like it, then it’s not good coffee for them. And if they like their toast a lot darker than you like it, it’s probably okay toast, but not good.

Your processes can be described as good as you can, but still the result might suck. Please analyze your processes and understand what quality aspects and decisions can be made at every step, and how they influence the outcome of the product. And learn how to understand the actual needs and wants of your consumers to help create a product they like and evaluate as “good”.

2 thoughts on “Context eats Process for Breakfast”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s