Back in 2015, McKinsey published a paper on the future of risk management in the banking sector. At the time I sent it to many of my clients in other sectors because of its very important messages, including this one in their summary:
“Bank risk management will likely look dramatically different by 2025, when it has become a core part of banks’ strategic planning, a close collaborator with business heads, and a centre of excellence in analytics and de-biased decision making.”
The paper includes commentary on how advanced banks were digitising, using machine learning, for example to improve decision making in relation to loans and the potential for default. Since then, Australia has seen a Royal Commission into the finance sector with a well-documented focus on non-financial risk.
While a digital strategy is important to any organisation if you want to de-bias decision making, what about the bias in decision making where there is a lack of data? The answer is to test underlying assumptions by forming a hypothesis and testing that hypothesis by collecting data. And the data you need may not be far away.
An example might be that you have experienced high turnover in a particular professional skill set in your industry with salaries rising steeply and the ability to even find staff falling just as steeply. The prevailing view is that the cause of the problem is a lack of new graduates over recent years. A fair assumption. But only an assumption. It could also be that graduates are in demand in new industries for example.
This would be relatively easy to test by obtaining the numbers of graduates from universities over the past 5 years. There may also be data about where those graduates have ended up and there might be correlations between which industries/companies are presenting at the graduate employment career fairs. Knowing what the cause is gives you a better understanding of how long and how severe the lack of graduates may be. And potentially what to do about it.
All this without using deeper research skills employing statistical methods. Depending on the scale of your problem, deeper research may well be worth it.
Stay safe and adapt – with better measurement!