March 25, 2020
COVID-19 and Need for Decision Intelligence
by Craig Bedell and Ryan Trollip
The importance of decision intelligence is peaking. COVID-19 shows how powerful the skill can be.
The amount of vim, vigor and rigor brought to decision-making will determine the extent to which the insurance industry will harness the insights provided from enhanced analytics, artificial intelligence, machine learning and cognitive computing. The topic of operational decision management (ODM) is not new, but its importance is peaking as the topic of decision intelligence grows.
There still seems to be a gap in understanding among insurance experts about the straightforward topic of decision-making. Whether it’s for a macro/strategic decision or repetitive business process decision-making, there’s a clear, proven approach.
A recent article, nCov-19, Mitigate the impact to your business (Part 1), written by Ryan Trollip, does a superb job of walking through the decision-making processes using the current COVID-19 situation as a case in point.
Key Points in the Article
Metrics — When looking to improve decision-making in an organization, we need to identify what metrics we are trying to improve and then select decisions that have the biggest impact on those metrics. With the virus outbreak, understanding and monitoring key metrics using dashboards like the wildly popular one that John Hopkins made available, have been key in monitoring trends and understanding the impacts of policy decisions.
Subject Matter Experts and Decision Modeling — Leveraging knowledge effectively and not just data, is the key to delivering ROI quickly. Your expert’s business knowledge can be extremely valuable if you have the disciplines in place to effectively elicit this knowledge and represent it in a way that can be leveraged by others. We call this discipline decision modeling. Whether you automate those decisions or not, or whether the decision leverages machine learning or simply conditional logic (rules), the first step is to leverage existing knowledge (not data, yet) to break down the decision to understand its dependencies, understand what types of decision-making and data will be required to drive the decision. Again, this is where subject matter experts play a critical role.
See also: How Coronavirus Is Cutting Connections
An example of this, from the referenced article, is from an epidemiologist on how to mitigate workplace risk, looking at the density of the workspace, how regularly the area is cleaned, ventilation type used, supplies available, high traffic area touch mitigation. etc:
On the other side of the equation, we want to balance risk aversion against impact to productivity. Each organization is different, but generally it’s not critical to have all employees in the office all the time. Even for companies that have no critical, in-office needs, it is clearly preferable, for productivity and culture, to have some roles in-office over others — e.g., a design team that needs to white board and discuss ideas vs programmers executing on a design.
There are other factors. For instance, some employees are more self-motivated than others to work without in-office supervision, and there are varying levels of technical ability to work remotely. If we take this day by day, role by role, employee by employee, for many companies, it would not be critical to have each employee in-office every day. If we rank the criticality of these in-office days and chart them, employees ranked by criticality (y axis) against level of in-office criticality per employee per day (x axis), for many companies, it would look something like the chart above. The intersection on the chart (red lines) is a rough illustration of how risk will likely fall faster than the impact to the business when reducing the in-office employees for many businesses.
Besides providing some insights on COVID-19 risk management decisions, we hope this article illustrates the process of crafting decisions.
The use of insights produced through new technologies, data and systems will only be valuable when it affects the main line decision-making processes across the insurance industry and in particular in the underwriting, claims and policy/customer administration processes.
Developing decision intelligence and decision management competencies is like any other skill. They need to be taught/learned, practiced and matured. There’s no once-and-done. Sure, there’s great technology to assist, but the business of insurance needs to evolve its decision making.
See also: Coronavirus: What Should Insurers Do?
Whether you are targeting macro decisions like the one illustrated in this piece, or day-to-day operational decisions – pick a straightforward place to start. Lead it with an excited line-of-business sponsor. Select a team of willing business process practitioners and engage in a decision-making workshop or design thinking workshop to uncover imaginative and informed ways to hone the decisions at hand.
Then you will be in a position to engage with operations and technical folks to examine how to institutionalize these new approaches.
Don’t forget to measure the results so that the outcomes are within expected tolerances and that all lessons learned are captured for the next advances in these newfound capabilities.