For our annual customer conference in March, we overhauled our agenda to focus on the planning challenges posed by the global COVID-19 pandemic. This year, customer presentations were rife with language like “commercial overrides,” “exceptional demand filtering” and “dummy demand.” All of these terms describe emergency exceptions that require an optimal symbiosis between humans and automated planning systems.

“Optimal symbiosis” sounds like something everyone would want — but what does it actually take to achieve it?

S&OP maturity. Our customers would almost unanimously agree that on the human side, it requires collaborative, cross functional planning and decision-making. Without exception, those that have achieved that symbiotic relationship have put processes like sales and operations planning (S&OP, or variants such as IBP, or SIOP, or SOE) in place.

Having an established S&OP process is particularly important during a crisis. When tensions are running high, personal agendas, biases and time pressures can skew the judgment of a single person’s or a siloed team’s decision-making. By comparison, the checks and balances inherent in the S&OP process help to generate plans that are reliable, sound and produce optimal business outcomes.

However, to manage successfully through a crisis period, you need to go well beyond what Gartner defines as Stage 2 maturity. It’s essentially an “inside-out” process of gathering historical data from different sources, consolidating it into a spreadsheet, and holding a monthly meeting to agree on the operational supply plan. But to best cope with an unfolding and unprecedented situation like COVID-19, you can’t be supply-driven or rely on history. You need to be at least at Stage 4, with an “outside-in” process of being able to read and learn from demand signals to make accurate predictions.

To deal with the exceptions that arise in crises, you also need to get a wide variety of stakeholders involved in S&OP. Bringing in marketing, for example, creates opportunities to introduce demand-shaping activities lsuch as promotions to deal with excess inventory situations. Involving new-product development (NPD) helps the team decide whether to launch new items to support customers in a crisis, and also gauge the impact of these on demand, strategic sourcing and operational and logistical capacity.

Data maturity. Before addressing systems, let’s get down to basics and talk about data hygiene. If you go into a crisis with dirty data that’s biased, inaccurate, improperly segmented, or in the wrong format, you’ll never get ahead of the situation. There are simply too many complexities to deal with once a crisis is underway. Here are just a few scenarios that our customers have faced during COVID-19, all of which rely on good data hygiene to guide decision-making:

• Planning for perishable foods and medicines that need cold transit and storage.
• How to serve all customers when some are hoarding — or more likely, buying “a little extra.”
• Minimizing obsolescence and waste resulting from bottlenecks in a just-in-time supply chain.
• Supply shortages of “far-sourced” materials and finished goods.
• Changes to the distribution network and sales channels; for example, using closed retail outlets or “dark stores” as mini-warehouses to cope with the surge in online sales.
• How to plan for returning to “normal.”

In a crisis, you really see the benefits of having a wide range of data sources, segmented by demand signal and deviation pattern, being factored into your planning. This involves extending beyond traditional enterprise data sources to include external, causal information like weather forecasts, customer sentiment in social media feeds, and location data. The more diverse data sources you can integrate into your planning, the more accurate your predictions will be.

Planning systems. Finally, if you’re going to factor in diverse and big data, you definitely can’t rely on spreadsheets and old-fashioned planning systems that draw on historical data, and use dated and simplistic methodologies such as weighted moving averages. These systems aren’t fit for the complexities of modern business, never mind a crisis. It’s time to put them in a museum!

The holy grail of S&OP is Stage 5, which according to Gartner includes “a move toward algorithmic SCP [supply chain planning] with appropriate automation of key elements of decision-making.” We have long advocated that forecasting be as automated as possible, so that people involved in planning are freed up to apply contextual knowledge of exceptional and changing circumstances — in some cases, to override and adjust the forecasts.

With this extra freed-up time, why not take advantage of one of the biggest powers of S&OP: the use of scenario plans to create “what-if” models of various demand and supply disruptions? That way, when the data starts suggesting you’re drifting toward one of those scenarios, you’ve already worked out a cross-functional plan. This allows you to react more quickly to changes in the market, and ideally gain competitive advantage from the more rapid response.

The best software to support S&OP employs “stochastic” or probability forecasting methods, which work like weather forecasting systems, providing a range of possible outcomes, weighted around the one that is most likely.

Following are some exception COVID-19 planning stories.

Ulabox: meeting escalating demand for home food delivery. For some companies, COVID-19 brought surges in demand. Ulabox, the first purely online supermarket in Spain, reacted efficiently to exceptional demand spikes. The company was able to understand the impact of changes on shipping times, capacity constraints, shelf-life concerns and other factors in its multi-echelon environment. By changing one thing — in this case, the forecast —  the system understands how to stock and where problems might arise. This “commercial override” capability helps Ulabox achieve the target service level for all its “A-class,” even in crisis conditions.

Franke: collaborating to better understand demand. Franke, a kitchen equipment manufacturer based in Switzerland, uses advanced planning software in more than 50 global regions. To support its S&OP process, it draws on various capabilities to minimize supply-chain disruption, including commercial overrides, exceptional demand filtering, adding dummy demand, and factoring in the partial closure of sales calendars. These functions have enabled Franke to generate a more accurate forecast amid extreme demand changes.

These adjusted forecast values are shared via a “collaboration hub,” where regional markets have two days to fine-tune the adjusted forecasts for their regions before the commercial overrides are imported back to the planning system.

McDonald’s Mesoamérica: meeting restaurant challenges. The impact of COVID-19 on restaurants has been significant. For McDonald’s Mesoamérica, the primary challenges are reduction in volume and forecast uncertainty. Its team also has to meet exceptionally high service levels while managing perishable products. The company can see all inventory by warehouse, location and item in near-real time. It can also consolidate by region for categories, whether beef, dairy or bakery products; in the south or north, or in a country or a single warehouse.

The team is currently working on short-term forecasting. Instead of relying on four years of information, they use two days of information to model the realities of COVID-19. These tools have made the company highly efficient in making key decisions.

Advancing the maturity of your S&OP, data and planning software isn’t just for managing through crises. It will make your company resilient, competitive and ready for anything — including being able to take advantage of, and profit from, positive market changes.

David Barton is general manager for North America with ToolsGroup.