“Magic 8 Ball, what will my demand forecast be in April 2021?” With a few shakes, the answer appears, “Concentrate and ask again.” A few more shakes and then it says, “Reply hazy, try again,” and finally, “Cannot predict now.” Of course, a mass-produced toy can’t consistently predict demand, but even with more sophisticated tools and processes, supply chain planning may not be able to make sense of the “historical” data available a year after a global pandemic.
The circumstances of early 2020 created some of the most extreme changes to demand patterns in history and there is no single solution for what to do going forward. For example, there will be a need to both depress future forecasts for household supplies — one of the few categories to see increased consumption — and lift demand curves for everything else. Factor in other characteristics, including no indication that all possible solutions have been found and no definitive stopping rules, and demand planning fits the definition of what design theorist Horst Rittel called a wicked problem.
This is the type of problem that is difficult to define and seemingly unsolvable because of the many interdependent factors and stakeholders. Supply chain planners will argue that this is the status quo, but the combination of complexities in 2020 should act as the final warning that “the way that it has always been done” will be inadequate in 2021 and beyond.
The urgency to revolutionize planning is understood and growing
The looming impact of inadequate planning processes is echoed by the May 2020 IDC whitepaper "Supply chain planning drives better business performance." Author Simon Ellis highlights that the volume of companies that view supply chain planning as a competitive advantage will fall from 65% to 35% within the next three years, suggesting that, “Companies either do not view their supply chains as adaptable or are unconvinced that the resources they have in place today will service them well into the future.”
With this anticipated drop-off in mind, management is most often placing planning as a top three transformation priority for both the next 12 months and three years. However, is an even faster response needed? And will those who don’t act have a chance to recover?
In a recent Bloomberg article discussing the work of Michigan State University professor Steven Melnyk, the deadline for developing an actionable plan is said to be much sooner, “by October,” to be exact. Melnyk further explains, “The winners will be those firms who are willing to make investments now — when demand is down — in preparation for future growth. Once the economy recovers, firms will not be able to make these changes because they will be too busy meeting demand.”
Solving wicked problems requires coordinated shifts in thinking and technology
The urgency is growing, and acting on changing planning approaches still requires solving a wicked problem. Over time, design thinking and advancing technologies, such as digital twins and artificial intelligence (AI), have been suggested as possible solutions. Neither seems to have been singularly deployed yet in a way that completely solves the complexity of demand planning, but the intersection of strategy and technology has greater potential.
For instance, Sandy Pentland of MIT said that despite limited instances of understanding causal structure, AI can act as a complement to the skills of people. He addressed this in a conversation with Deloitte Insights, saying “A middle-class chess player with a middle-class machine beats the best chess machine and the best chess human. I think we see a lot of examples coming up where the human does the strategy, the machine does the tactics, and when you put them together, you get a world-beater.”
In practice, intelligent digital twins provide a real-life sandbox in which the design thinking phases of ideate and prototype can be worked continuously. The information that is gleaned from those steps can then be reintroduced into the process, allowing for an ongoing evaluation of possible changes in approach.
What that means for supply chain planning is that shifts in thinking and new tools must be concurrently deployed as part of an iterative learning cycle that treats forecasts almost as a living organism. As AI leader Rubikloud describes it, “Any effort to accurately forecast demand amidst rapidly-changing business complexity requires not only complete data, but the ability for the platform to automatically scale as business increases, for employees to augment or override insights when needed and for the data to reintegrate back into legacy systems.”
Without that type of blended investment, efforts to create the perfect forecast will continue to feel like using a Magic 8 Ball. Questions will be asked repeatedly without a clear sense of progress and the wicked problem will live on. By that time, the window to act will have closed.
Instead, see 2020 as the ultimate catalyst for change. There is an urgent need to reset expectations and establish new ways of thinking. That, when blended with advancing technologies, will create world-beating supply chain planning.