I’ve written before how risk aversion encourages predictive project management. This morning I was passed a link to Ryan Moulton’s blog in which he explained mathematically how risk-reward calculations leads to this behaviour. Here I explore the questions he raised.
Why are big organisations so risk-averse?
Big organizations ‘play’ with huge amounts of money and have many people dependent on them, so they have to ensure that any investment they make is wise.
As individuals we have many options to choose from when we decide what we do with our own money: we can keep it under our mattress, where it will progressively decrease in value (relative to what it can buy over time); we can invest it in a (traditionally) safe option like a normal savings account, where it will attract interest, but often barely about the inflation rate; so we know that to get significant returns on our money we have to invest it in options that have less certainty about them, i.e. may go down in value as well as up, but when they do go up, boy they can fly.
Different people have different attitudes to risk, which is why most financial firms have a range of investment types to suit all tastes, many offering ways of balancing a client’s money across several categories, from safe to risky.
Now, when this is multiplied up, a big organisation has many people, each with their own attitudes to risk; when you combine that with the fact they’re dealing with other people’s money, you often end up with an approach that suits the lowest common denominator, i.e. the most risk-averse will likely take precedence.
How do big organisations seek to avoid risk?
Once we accept that organisations are going to be risk-averse, we then have to consider what behaviours this generates.
Thinking back to a personal example again: when we buy a car, if we are risk-averse we might decide to spend ages thinking about it, reviewing write-ups about cars, thinking carefully about what we’ll need the car for (family, dogs, towing, sports equipment, etc. etc.), shortlist some; then maybe we’d sleep on it, for a while, for a long while, before finally taking the plunge. This is because we’re really not sure if we’ve thought about everything we need, whether we know all the options, and we want to avoid buyers remorse, that feeling the next morning when you pull out of your drive and see another car that just looks like it would be so much better!
Now again, multiply that up, in a big organisation, where the most risk-averse will hold sway, that same careful step-by-step approach makes so much sense doesn’t it? How can people think of committing all their resources and money onto a project before they’re really sure that they know exactly what they want it to do, look like, etc.
Or does it? Is there another way?
Understanding how personal attitude to risk is magnified and distorted when you scale it up goes some way to explaining why many big organisations still adhere to a largely predictive and plan-driven approach to minimise risk; because on the individual level it is common sense that people should nail what they want to do before they start; when you scale that up to the organisation level, wanting to avoid risk at all costs often drives an uber-waterfall approach.
There is another way
If these typical risk-averse behaviours led to perfect results, eventually, then although they have to wait longer, it would be worth it. However, with so many projects still failing to deliver promised benefits, on time, to cost, and of consistent quality: something is clearly wrong with this picture.
When people spend so long determining what they might do, they miss launch dates and so lose market advantage. Sometimes to keep to the launch dates they compromise on the technical strength or the business value and so miss out on the opportunity by delivering something that is not what customers want or has so much technical debt it drives a poorer customer experience.
There is strong evidence that taking a more adaptive and value-driven approach works to minimise risk in a different way; by ensuring that they know just enough at any one time to deliver a small increment of what they want—with each increment being free of technical debt (*1)—they can deliver to market as soon as they have enough increments to give enough value and build on that.
While some may feel that this means dropping half-built products into the market, in many ways organisations already do that when they launch products with manual work-arounds; however, this is more about frequently dropping smaller increments of products that work and on which they can deliver enhancements in progressive releases.
footnote *1: for anything to be truely ‘free of technical debt’ of course it also has to be architecturally sound, which means a balance of prioritising good architecture alongside important business needs