A triad for success
“work backwards, make small changes, get fast feedback”
(This is the second in a series on change and software engineering. While this post is quite general, my earlier post and this one set the stage for a more technical discussion in later posts)
The world around us seems to be changing ever more rapidly, and it is unlikely we have the knowledge, the skills, or the resources beforehand, to succeed at our ventures. We need not despair: we can iterate rapidly, and move confidently towards pretty much any goal using this “triad”, as I shall illustrate below.
Working backwards; we begin, not at “the beginning”, but at the end. We assume we have succeeded at our goal: we then visualise the world in this end state, and attempt to continually trace as many paths back as we can, to where we currently are. At all times, we create solutions that contain the essential elements of the end-state, as soon as possible.
The sculpture is already complete within the marble block, before I start my work. It is already there, I just have to chisel away the superfluous material. — Michelangelo
When working backwards, we get to our goal (like Michelangelo) by revealing more and more of the form we aspire to.
To work backwards, I must approach my goal with humility and a sense of wonder, thankful for the opportunity to do things differently, and to learn afresh. I seek to avoid entering the dreaded “spiral of familiarity”
The spiral of familiarity begins with comfortable assumptions, rushes forward with current knowledge and skills, then plunges towards foregone conclusions, and possible despair.
When not working backwards, we tend to chart a way forward in terms of what we already know. We are likely to make many assumptions, including several that we are not even consciously aware of. Our current knowledge may be inadequate for the purpose, and can lead us astray. We are making our way through a fog, in a direction illuminated by the light of past experience, and we may lose opportunities to innovate.
Lastly, working backwards reveals the unknowns we must tackle head-on. Otherwise, we may wallow in a false sense of comfort by taking on just the familiar bits, and it may be too late, or we may consume other scarce resources before we realise we may not get to our goal.
Make small changes: As human beings, we have a limited ability to comprehend complexity, say, when presented with many moving parts (see “The magical number seven, plus or minus two: Some limits on our capacity for processing information”). We model the world around us in terms of simpler abstractions to conquer complexity. Each such model involves multiple “axes” of change. Our experiments consist of making a change along one axis at a time, while retaining others (the “controls”) unchanged; and our ability to predict the outcome rests on many such experiments.
We should therefore, organise our efforts, so that we can, quickly and easily, and at all times, make a small change and observe the effect.
In our impatience, we may be tempted to organise things so we can either only make relatively “large” changes, or we can only make changes spaced out at particular intervals in time. It may not be immediately obvious, but these “evil twins” travel together: where one of them is to be found, the other cannot be far behind!
Get fast feedback. We should strive to create systems that give us feedback instantaneously. Without fast feedback, our ability to predict outcomes suffers immensely, and we must gamble!
The lack of fast feedback inhibits innovation in many ways. The areas of human endeavour that make the slowest advances, as well as the ones that demand the most upfront and risky investment, all provide slow feedback. Democracies require years to (in)validate electoral choices. Innovating in aviation requires gambles that consume decades, and breeds systemic failures. New medicine ordinarily requires trials over years.
In the examples I have cited above, the lack of fast feedback may be “structural” due to several reasons. Safety (and consequently, regulation) is paramount in both aviation and medicine. However, COVID-19 has galvanised humanity into working together to get fast feedback on potential vaccines, in just weeks or months. Perhaps when the crisis has passed, we will have learnt more enduring ways to get new medicine out both safer and faster!
Small changes, over and over, and faster!
Our ability to both make a small change, and immediately observe its effects, helps us conquer complexity. We seek to go forwards, and backwards as we wish through a stream of changes, such that the changes we make are reversible.
Thus, having learnt from fast feedback, we begin again at the first leg. Working backwards, we identify and promote the next small change, and learn from fast feedback, over and over.
We strive to make the change in each cycle as fine-grained as we can (the “quantum of change” shrinks). We also seek to increase how often we can repeat this cycle (the “cadence of change” becomes more rapid)
It is common experience to be confronted by “big decisions”: when our choices involve high stakes. However, if we can make, as well as reverse any number of small changes, many such “decisions” would dissolve!
I consider the triad liberating for a number of reasons:
- We need not be paralysed by challenges seemingly beyond our current ability to comprehend
- We need not make many decisions with high stakes
- We adopt a creative approach at all times, to discover, learn, and improve!
In future posts, we will visit the state-of-the-art in software engineering to both make systems easy to change, and to empower ourselves to follow the triad for success.
