6 Steps to Predicting Shifting Demand Patterns While Navigating the Coronavirus Crisis
We have not experienced a global pandemic like the coronavirus in last 100 years. The sheer increase in demand for everyday necessities like toilet paper, sanitary wipes and bottled water is putting undo stress on a lean global supply chain.
It is testing the agility of many retailers and consumer packaged goods (CPG) companies as they attempt to ramp up manufacturing facilities and logistical operations while struggling to keep up with consumer demand.
Business executives are looking to data, analytics and technology for answers on how to predict and plan for the surge and, ultimately, the decline in consumer demand. It is significantly easier to shut down facilities than it is to quickly boost production and capacity.
The biggest unknown is whether there will be a delayed economic recovery or a prolonged contraction. Regardless of the outcome, retailers and their CPG suppliers will need to think ahead and be prepared to act quickly.
How to predict and plan for the surge and decline in consumer demand patterns
Retailers and their CPG suppliers are the backbone of the consumer goods supply chain and a lifeline to their customers. Their ability to operate efficiently is determined by the weakest link in the end-to-end supply chain.
The current crisis has changed the make-up of the average grocery basket making it difficult to predict rapidly changing demand patterns. As a result, the current supply chain is struggling to keep up. Restoring balance will require changes in the way demand forecasting and planning are conducted by both retailers and CPG companies.
Navigating the current climate will require new intelligence, resilience and more dependence on advanced analytics and machine learning than ever. Here are six actions that can improve retailers’ and consumer goods suppliers’ ability to predict the changing demand patterns.
1. Use downstream data that reflects true consumer demand
First and foremost, analyze and forecast the POS data collected through store scanners. Use it to determine the shift in demand patterns for your products in order to more accurately forecast the mix within the average market basket.
It is even more important to focus on forecasting the lower product mix as it will indicate which items have the highest demand velocity as well as those products with the lowest demand. Most CPG companies receive POS data directly from their retail customers on a daily and/or a weekly basis and POS data is the truest consumer demand signal.
Several CPG companies have been analyzing and forecasting the POS data for products sold to their top 20 global grocery retail customers, are recognizing significant shifts in consumer demand patterns almost immediately and are acting accordingly. They have also detected the increase in demand for their products on Amazon.com as consumers shifted from brick and mortar stores to online purchases during this same period.
Using POS demand history and revised future forecasts as a leading indicator in their shipment models, CPG companies can more accurately predict supply replenishment to those same grocery retail customers.
This new consumption-based forecasting approach using the MTCA (multi-tiered causal analysis) process has allowed CPG companies to significantly improve not only shipment forecasts but to also detect turning points in demand patterns much faster than traditional shipment models allow. (See Chase, 2016, Next Generation Demand Management: People, Process, Analytics and Technology, John Wiley).
2. Adopt and implement advanced analytics and mL algorithms in your demand forecasting & planning
Implementing advanced analytics and machine learning algorithms can help spot abnormalities quickly and adjust immediately. Several SAS customers who have recently implemented advanced analytics and machine learning technology were able to predict the shifts in demand patterns quickly, while their legacy systems were failing to predict those changes.
Customer Example: Recently, we had a discussion with a large French retailer’s data scientist about how the brand was coping with the COVID-19 crisis, and how their new SAS solution (designed to forecast warehouse shipments) was handling the shift in demand patterns. The data scientist explained that the new forecasting was performing very well while their legacy solution was crashing.
This comparison illustrates that even when the shifts in demand patterns were significant, the autotuning advanced analytics models quickly adapted.
3. Implement a short-term demand forecasting and planning process
Implementing a short-term (one to eight weeks) forecast that utilizes advanced analytics and machine learning to predict weekly and daily demand using sales orders and shipments in combination with POS data.
Using POS data as a leading indicator in the models (along with sales promotions and events) allows a retailer to calculate not only the promotion lifts but the shifts (anomalies) in short-term demand patterns.
3. Incorporate social media information
Incorporating social media into your demand forecasting and planning process by capturing consumer sentiment. Text mining and sentiment analysis allow retailers to monitor social channels for consumers’ comments on product availability, what’s trending, and their store purchases.
Once you’ve gathered a large enough sample of customer conversations, you can apply sentiment analysis to determine which products are moving rapidly off store shelves, which ones are completely out of stock, as well as additional changes in purchase patterns and store availability.
Working closely with the marketing and/or consumer insights team demand, planners can utilize this information to identify in real-time the key stores, categories and products that are affected.
For example, paper goods are moving faster in the northeast regional store clusters in the NYC metro area versus the mid-Atlantic stores. So, the planner should focus demand planning efforts on the northeastern NYC metro area store clusters, categories and products. This information can be easily integrated into the demand planning process by simply working closely with the marketing and/or consumer insights team.
Sharing analytics findings and information across departments is vital to accurately predicting shifts in demand patterns. This is a core reason for embedding demand analysts and planners in the sales/marketing organization.
5. Focus on the granular view and regional geo areas
Patterns in consumer demand are varying across countries and product categories more so than usual. Many retailers are experiencing huge spikes across local geographies in excess of 800% for over-the-counter cold and flu medicines while food items are in excess of 25-50%.
According to a March 16, 2020 McKinsey & Company briefing (What food retailers should do during the coronavirus crisis), the change in consumer demand shifted dramatically in the periods before the Italy lockdown. Sales for cleaning and safety products like sanitizing alcohol, tissues, bleach, hand soap and toilet paper increased between 23% and 347%, while raw ingredients and long shelf-life products like flour, rice, pasta, pasta sauce, frozen food and water had lower increases: from 20% to 82%.
Meanwhile, discretionary products like sweets, baking mix, cosmetics, perfume, and salty snacks decreased anywhere from by 4% to 52%.
According to the McKinsey briefing, in some geographies, consumers were buying fruit over beer but, after a few days they were returning to beer and snacks as they found themselves at home for extended periods of time.
Subsequently, certain store formats like convenience stores are seeing huge declines in sales, while others like e-commerce are experiencing up to 700% increases in consumer demand and may be unable to fulfill customer orders.
Collaboration and full transparency between retailers and their CPG suppliers are crucial to identifying and acting upon demand signals and changes in demand patterns. Constant communication will enable retailers and CPG suppliers to act fast and appropriately to mitigate root cause threats that contribute to under-predicting demand for essential items.
6. How do you handle the abnormal historical data after everything goes back to normal?
Another consideration on the minds of many of our customers is the mid-term: when the Coronavirus crisis begins to subside, and demand returns to normal, how do we adjust the demand history? As the dust settles and we again see normal demand patterns emerge, there will be a need to address the abnormal demand patterns in the historical data.
The biggest challenge for demand analysts will be to cleanse those abnormal demand patterns out of the demand history to reflect normal demand patterns. Many will fall back to the practice of manually removing the abnormal historical demand without realizing they are erasing key information, as well as possibly over- or under-projecting what the normal demand would have been if the crisis had never occurred. This presents a valuable an opportunity to learn from a tragic situation.
The best approach is to view those historical abnormalities as outliers and agree not to manually cleanse the data. This is another opportunity to capture those outliers and adjust the historical demand using advanced analytics.
By simply adding outlier variables (also known as dummy variables or intervention variables) to existing models, the demand analyst will be able to capture the abnormal demand patterns whether positive or negative, as well as automatically optimize the historical demand to reflect normal demand patterns. More importantly, they will capture those patterns to be used in future crises. There will be no need to manually replace all or part of the abnormal demand historical data or input missing values for those dates.
In other words, let’s view this unprecedented crisis as a learning event. Using more advanced modeling techniques, we can capture the shape of the event and remove it from the history so it can be reused in the future if something similar happens. We hope it never will, but it’s a best practice to capture the impact, so you can easily add it back into the future to better predict the outcome sooner, rather than later.
The approaches outlined in this document offer a framework for thinking rigorously and systematically about how to forecast changing demand patterns during a time of uncertainty. These recommended actions should become part of your ongoing demand planning discipline. This will enable retail and consumer goods companies to judge which analytic tools and technology can — and can't — help them make real-time decisions at various levels of uncertainty.
This approach provides a playbook to tackle the most challenging decisions that demand analysts and planners face right now, offering a more complete and sophisticated understanding of the implications of capturing and predicting changing demand patterns.
This unprecedented pandemic is a wake-up call to all industries, including retailers and their CPG suppliers. It’s no longer just about collaborating across internal departments: It’s about humans partnering with machines in an autonomous supply chain with full transparency.
There is a mandate to leverage the collaboration between humans and machines to fight an invisible enemy that threatens our way of life and our economy like none other. Together, we can detect abnormalities faster, identify immediate shifts in demand patterns, and make decisions in real time. There’s no time for hesitation or mulling over options.
As executive industry consultant, Charles Chase is a thought leader and trusted adviser for delivering analytics-driven solutions to improve SAS customers supply chain efficiencies. Chase has more than 20 years of experience in the consumer-packaged goods industry, and is an expert in demand forecasting and planning, market response modeling, econometrics and supply chain management.
Chase is the author of two books--Next Generation Demand Management: People, Process, Analytics and Technology, and Demand-Driven Forecasting: A Structured Approach to Forecasting. He also co-authored a third book titled, Bricks Matter: The Role of Supply Chains in Building Market-Driven Differentiation. Follow him on LinkedIn. On Twitter, follow @SASsoftware and @SASRetail.