The Ambidexterity Continuum of Testing

I was introduced to Continuum as an English word many years ago when James Bach came up with Scripted/Exploratory Testing Continuum which I think predates Exploratory Testing 2.0/3.0. I am including it below for your reference:

Even today, if one were to really understand what is meant by these terms, this diagram IMO is the single most important visual available out there - simple yet powerful.

The word continuum itself had a huge impact on my way of thinking, and I started using it to explain many more concepts to myself & others. I use it in a lot of places today. It's a very powerful thinking tool in my opinion. This article is about one such continuum.

Here's another building block as a foundation of my testing thought process - The Ambidexterity Continuum of Testing as shown below:

Purpose

I believe in the highway of Testing. Everything else should amalgamate back into the highway so that the main body of knowledge gets enriched over time. As that does not always happen, I have tried to do that for myself. So, exploratory testing and scripted testing are good distinctions but not types of testing for me. I had this discussion with James in 2012, the only time I have met him in person and we were both ok with me thinking in this manner for myself. I had recited a Punjabi poetry to him on the concept of History as a traveller of Highways, which I guess I'll write sometime elsewhere :-).

I thought to come up with a continuum that explains other subtle aspects, which are integral to my way of thinking.

Why I find this continuum useful for myself is because all the three zones exist in me. I am a consultant, so I need to create a balance between what the problem is and who I am. As I have told earlier, I am not a perfectionist. I just seek balance in these zones. When I get it right, I am happy.

Exploration vs Exploitation

Exploration is about the uncharted territories of opportunities and possibilities. Humans have the primary role here, although they can guide tools to help them in parts of this exploration. Think innovation. Think unknowns.

Exploitation is about making the journey to already explored paths cheaper and less time consuming. This is achieved with documents, diagrams, models, tools, custom code, frameworks and so on. The idea is to reduce human involvement as far as possible or the level & type of human skills needed for the task. Think efficiency. Think known, well-defined stuff.

The Three Zones of Horizontal and Vertical Continuum

The diagram has a horizontal continuum of 3 zones, wherein each zone is also a vertical continuun.

These three zones that derive their names from the Japanese Shuhari concept are: Shu, Ha, Ri zones.

All zones co-exist to varying degrees in an individual or an organisation. What matters is what is the distribution of focus amongst these zones at a point in time and typically the distribution keeps changing (or rather must change) over time.

Let's look at them one by one:

Shu (Obey) Zone

This is the Stay Alive zone.

The purpose is to do good enough testing that we are not missing the fundamentals of what Quality means for a context. The goal is to move work into Exploitation zone as much and as quick as possible.

This is the zone to get repeatability right - you want low cost solutions for quick and short term gains.

For consistency, this is the zone where rules are the way and organisational needs are placed before individual needs because the zone targets the level of Quality which can not be compromised upon. It is not the zone for heroes. It is the zone of uniformity.

Innovation happens, but only in smaller incremental chunks to be absorbed into repeatability.

The work is deliberately targeted to be simplified and hence becomes mundane. Quick and cheap automation should be a primary goal so that for the same cost and time, exploration can still be accommodated.

This also happens to be the zone which suffers from the Pesticide Paradox because of the inherent repetition, which in my opinion is not a bad thing if this is not the only zone in you and/or your organisation.

Ha (Detach) Zone:

This is the Stay Relevant zone.

There are obvious limitations in the Shu zone, howsoever critical it might be. This is the playground of good testers as well as deeper automation & exploitation.

The one-to-one correspondence however does not exist. It means all ideas discovered using exploration can not be represented in the exploitation zone. You will always need testers who continue doing the exploration. You will have experimental automation along with deeply architected frameworks. You will have opportunities for advanced test design and pattern thinking.

The success factor is how these ideas can be translated and transported into the Shu zone to improve the quality of work.

While the organisation still reaps the benefit, this zone is tester-centric and accordingly the gains can be slower but worth the cost and effort it takes.

This zone is about long term impact on the Quality. While Shu zone is staying alive zone, Ha zone ensures that it remains that way as time goes by.

It is still thought of as investment with radical innovation as a typical expectation.

Ri (Leave) Zone

This is the Stay Malang zone.

Malang is word used in many languages potentially of Persian origin - which means one who does not care about anything. You can think Nirvana, which in my opinion is not a state of perfection rather a state of being unconditionally happy.

This is zone is completely about testers as individuals who don't pick up a concept because they want to solve a specific problem. They just do what they do because they love testing. They think finally all this is adding up in its own mysterious ways. They know they don't and can't know everything.

As can be understood, the cost and time of this is indeterminate. If you are an organisation, you can not create Malangs, although you can build a culture with this hope. As an individual as well, it is a difficult choice as we all are controlled by many responsibilities. So, this zone as a 100% zone is impossible.

Having said that, this is the zone of liberation. I have found some such partial Malangs in my career and I think they are the ones who leave a lasting impact beyond the boundaries of any given organisation, on the community as a whole.

This is the toughest zone in terms of expectations as it is about more open-ended innovations. It can be felt, not really measured.

Typical Problems Across the Zones

• Some (many/most?) organisations and testers tend to think as if they are forced to choose between manual testing vs test automation, scripted testing vs exploratory testing and so on. There is no versus here at all, it's all about how much of what you should be doing and are doing. I am recently seeing posts that are talking against regression testing (!!!). I'll write separately about the last one in detail on how misguided it is.
• 100% Coverage is not possible as it will need infinite tests and effort. But have you seen anyone reporting that we have 0.000000000000000001% coverage? No, will not make sense at all too. So, what is done? Types of coverages get defined. Till this point there is no problem, that's what we need to arrive at our version of affordable achievable quality. However, to sort of hide the fact that most testers are busy in exploitation part rather than exploration part, some extremely poor coverage metrics are chosen. E.g. test case execution coverage to indicate how many of the documented test cases were executed. One must understand that this is a poor metric even within the Shu zone. This is not just for black box testing, even on unit testing side, one can see why statement coverage is still widely used as a metric, although it can infact be manipulated by writing partial code blocks with poor code as a side-effect.
• Testing is both an art and a craft, as widely acknowledged. Can you be an artist for whole day, every day producing pieces of art? There are also testers who overplay the Ha and Ri zones without realising how impractical that will be and that there is value in Exploitation parts of this continuum as well. In my opinion, humans as explorers should feel happy when they are able to convert some of their explored ideas into the exploitation zone. It frees them and it makes the already discovered ideas cheaper to execute.
• There is still another problem which I have seen. On one side we have testers picking up random information, code from the web and making it a part of their work and on another side perfectionists who find flaws in every approach. I think we need to seek some balance. We need to start somewhere from a solution perspective, which will obviously be not a perfect one. In this spirit though, we should not be starting with something which is completely wrong and do regular refactoring (I'm using this word in a generic sense. Documents can be refactored too). That's the reason, it's absolutely critical to know the basics, which are sometimes lost in all the noise.

That's all for now.

Some related articles:

20 Years of Contradictions

Pluralism and the Infinite Testing Schools

The Infinity Model of Imperfect Quality for Fallible Testers

An Undefinition - What is a Test?

AEIOU - The Vowel Model of Thinking for Pluralistic Testing

Handling Complexity in Testing - What's Your Slice?

Uncaging, the Perpetual Translation Engine and the Deltas