Q) During the current pandemic we are facing a very difficult time in preparing forecasts. Our forecast accuracy is far below what used to be. Can you suggest any way to improve it?

A.) We are certainly in a new economic phase, something we have never experienced before. In the past we had disruptions either in supply or demand—not in both as we are currently experiencing. This may be short-lived but we must make sure we deal with it. This means we need to change the way we forecast. Firstly, we should keep in mind that the sharp increases or decreases in sales data are not outliers but a reflection of new data patterns. When an outlier repeats itself again and again, it is no longer an outlier, but a part of new pattern. This means that old data is not relevant for future forecasts. Secondly, you need to know how the data pattern is changing. The data pattern of many products has drastically changed and the sooner we learn about it, the better. To learn about the change in patterns and to respond quickly enough, we need to work with not monthly or weekly data but with daily data. Compute the percentage change in cumulative sales from one day to the next, and then compute the average weekly change. If the weekly percentage change is rising, it means that the trend is upward; if it is falling, it is downward. We can use this trend to make a forecast for the next period. It may not be long before the pandemic is over. With that, the pattern will change again. The weekly percentage change in sales will quickly tell us which way the data is trending, and how strong it is.

I hope this helps.

​Happy forecasting!

Dr. Chaman. L. Jain,

Editor-in-Chief,

Journal of Business Forecasting

IBF recently asked demand planning leaders how they are guiding their companies through the pandemic. The following reveals the key insights from those conversations.

What Are Planning Leaders Doing Differently To Combat Covid-19?

Many companies have rushed to establish new, collaborative processes to understand the evolving demand and supply picture, and how to best respond from a supply perspective. In some cases, companies with mature planning organizations have realized that these processes are already in place, Camila Sierra, Sr. Director, Global Planning at Converse, commented, “We realized very fast that those forums that we were trying to create were just the S&OP forums we already had, and that all we needed was the right people round the table.”

“We realized very fast that those forums that we were trying to create were just the S&OP forums we already had”

An existing S&OP process has also proved valuable at Orchard Therapeutics, a global leader in gene therapy. Cristian Circiumaru, Associate Director, Global Supply Chain, revealed that before Covid-19, Materials Management and Inventory Management weren’t an issue – but they soon became one as supply suddenly became in doubt. They expanded S&OP to include Materials Management and Inventory management, as well segmentation of suppliers to better manage sourcing.

In Times Of Crisis, Get Back To Basics…

We may be embarking on the age of predictive analytics and big data, but a key theme for the planning leaders we spoke to is to recognize the limitations of new technologies when the inputs no longer make sense. “My biggest lesson learned last is to get back to basics”, observed Circiumaru. “Put the focus more on quantitative insights in demand planning rather than relying on sophisticated mathematical models. Yes, we have the technology now to support new models and those are in play as we speak, but refining qualitative insight is very important in 2021 to drive new changes in the forecast models”.

“My biggest lesson learned last is to get back to basics”

…Or Speed Up Innovation

Camila Sierra is currently drawing up plans for Converse to update their forecasting and planning systems, with Covid-19 having exposed the weaknesses of legacy systems and traditional modes of working. “We’re investigating automation because our teams are overworked, partly because they’re using too many spreadsheets”, she remarked. She continued, “We’re looking how to use technology to drive S&OP and create scenarios for more informed decisions. We’re creating an investment plan for the next couple of years to improve this area.”

“We’re investigating automation because our teams are overworked…too many spreadsheets”

A Temporary Return To Supply Driven Planning?

We asked Wallace DeMent, Sr. Demand Planning Manager at Pepsi Bottling Ventures how he is reacting to Covid-19 disruption, “I’ve been with the company 42 years now, but I haven’t seen anything like we’ve experienced with this pandemic. We’re used to basing forecasting financial plans and sales demand plans on what we thought the consumer would buy. We had a rude awakening having to base plans on what we are allocated from raw materials suppliers. It was a definite a paradigm shift in how we do business”.

“We had a rude awakening having to base plans on what we are allocated from raw materials suppliers”

Of course, S&OP doesn’t fix the supply shortages many companies are experiencing, but it can help companies work around them. At Pepsi Bottling, global aluminum shortages mean they must limit production of many of their core offerings – drinks cans. S&OP allows DeMent and his team react to a weekly allocation of raw materials from suppliers in a timely fashion. A weekly S&OP meeting adds value by communicating to Production what materials are available straight away. DeMent says that daily meetings are sometimes necessary to communicate changes, with early morning and late-night meetings currently the norm.

Combining S&OP & S&OE To Make Supply Chains More Agile

Circiumaru says that at orchard Therapeutics, they are relying on projections from their commercial and business development teams, employing forecast techniques based on population data and prevalence of diseases. He told us that the key is having a ‘control tower’ for supply chain, “As long as you have a solid foundation for S&OP, you have visibility into the entire supply and demand picture. We’re moving to a weekly S&OE process, which complements the full S&OP cycle. For Circiumaru’s therapeutics business, this visibility into demand and supply is not only desirable but absolutely critical, “Our service levels need to be 100% otherwise patients die.”

“Our service levels need to be 100% otherwise patients die”

Creating this control tower needn’t be complicated. Relatively simple and affordable tools can provide the much-needed visibility. “It can be as simple as, say, Tableau”, Circiumaru observes. “Someone who knows Tableau can plug in a bunch of data sources and spit out meaningful insights”.

Scenario Planning When You Cannot Forecast 6 Months Ahead

Forecasting is relatively straightforward when demand variability is stable. But when demand is highly volatile, new methods must be employed. Camila Sierra observed that at Converse “It’s been a challenge not being able to use any historical trends. We’re looking more at the last 90 days and where are our consumer buying. We also have to trust our senior leaders in terms of their bets on where the market will be in 6-12 months. Nobody has a crystal ball, but we need to make decisions. What has helped us is setting priorities like protecting margin vs revenue. That has helped us build a couple of scenarios that we can plan around.”

Cristian Circiumaru echoed the need for this kind of scenario planning to build demand plans around business priorities, “You need to take the principles of CPFR and expand them to Marketing, Brand teams, Product Development and Packaging teams. You want to have conversations where you explain different scenarios and their impact on cost-to-serve in terms of packaging and marketing etcetera.”

“Have conversations where you explain different scenarios and their impact on cost-to-serve”

For companies with existing S&OP processes, this is bread and butter; those without will have to scramble to establish the necessary collaborative forums.

KPIS In Times of Crisis

It’s not just processes that have to change when disaster strikes – performance metrics must change to. KPIs don’t always track what’s really going on in the business right now, says Wallace DeMent, “You have to be careful with your KPIs during Covid-19. Right now, some our forecast accuracy looks really amazing, but it’s really easy to get high forecast accuracy when you’re on allocation and you know exactly what you can sell! You have to go back a year prior to see what could have been sold, and use those findings to build the new business plan”.

“You have to be careful with your KPIs during Covid-19″

Of course, it’s not easy updating KPIs as market the environment changes. “Measuring S&OP is a tricky one”, says Cristian Circiumaru. “I have an S&OP scoreboard that I present monthly at the Executive S&OP meeting so all functions can see their own attendance and are held accountable. I have inventory min and max tracking for key materials as well as forecast accuracy, but not forecast bias at this point in time. Right now, making sure the right people are present at the S&OP meeting is the priority”. Having such KPIs established during normal market conditions is standard practice – having them in times like these is an absolute necessity.

Parting Thoughts

As vaccines are rolled out globally and lockdowns restrictions ease, the end is in sight. But, as Wallace DeMent observes, the challenges are far from over, “We still have a lot pandemic to get through which means more data having to be cleansed at a later date. My fun has just begun.”

Even before the pandemic, foward-thinking retailers leveraged AI and machine learning technology (ML) for demand forecasting. In the post-Covid world, however, those ML models have failed to provide accurate predictions because they don’t know that the data they use is now obsolete due to changing demand patterns. How can we upgrade them to the new reality?

There are six possible ways to get a more accurate forecast:

1 – Gathering data on new market behavior: As dynamics within the new market stabilize, use that data set to create a new model for forecasting demand.

2- Use a feature engineering approach: Track external data sources like price indices, market states, latest news developments, exchange rates, and related financial/economic factors. Using these, models can generate more accurate predictive outputs.

3 – Factor in up-to-date POS data: Analyzing recent POS data can allow us to observe and react to real-time shifts in patterns of demand, improving forecast reliability. Depending on a given product’s classification, the appropriate range for a POS data set might be between a month to two months.

4 – Use the transfer learning approach: If we possess any data sets relating to historical pandemics or behavior based on similar principles, we’re able to use that data within the context of this present-day pandemic.

5 – Utilize a model for information cascades: Merge the cascade modeling with current POS data sets to create a demand forecasting model that is able to recognize aggregated consumer behavior patterns and predict herd patterns for future sales.

6 – Leverage NLP (natural language processing) technology: NLP analyzes actual consumer comments and posts from an array of social media sources, from media platforms to popular social media sites. NLP can use sentiment analysis algorithms to collect and analyze conversations and discussions from real customers. This gives an unfiltered look at consumers’ behavioral patterns, preferences and attitudes.

If you are looking for a way to improve your current ML models and thinking of building a demand forecasting feature from scratch, this will help you to choose the best approach depending on your business type.

A data scientist generally works with historical data, and it’s impossible to predict such drastic changes as a worldwide pandemic. But as a general rule, you should prioritize flexibility in retraining your models, add more external factors as predictors, and account for a short-term perspective as long-term models become less relevant.

We tackled this problem with pre-Covid models for a restaurant business. Here is an example from our dashboard:

Forecasting dashboard showing normal revenue before the pandemic.

Forecasting dashboard revealing revenue post-lockdown.

In this case, we rebuild our models from scratch, losing a degree of accuracy. At this point, the historical data are not relevant anymore and we wait for new statistics and patterns inside the data.

Not every demand forecasting article on the internet fits the particular needs of your business and industry. Effective approaches vary dramatically depending on business types – here are some of the distinctions and varying ways to address demand forecasting:

Small vs Large Businesses

Small and large businesses should be approaching forecasting in completely different manners. First of all, the acceptance criteria for huge businesses is significantly smoother than for small businesses. We can have a higher error for prediction in quantity, as high sales volume allows for greater tolerance for error. When it comes to historical data, large businesses have a higher volume of collected data, making it easier to identify patterns in customer behavior. With small businesses, it’s often necessary to test your hypothesis to prove the correlations between sales volume and predictors.

Demand for a product at a large business, showing clear demand patterns.

Demand for a small business showing erratic sales patterns.

Online vs Offline

Online sales allow for a greater range of predictors and external factors in your modeling. It’s not necessary to have a POS (point of sale) system, as you’re able to collect all relevant data from the website. You know more information about customers and collect their historical purchases. With different predictors, you can apply a wider range of machine learning models: Gradient Boosting, Random Forest, SVR, Multiple Regression, KNN, etc. With offline sales, you are often limited to historical sales only. The best approach here is to use Time Series Analysis.

The USA vs Europe

Regional differences play a huge factor in predictive modeling, as varying locations will have different behaviors both in specific sales and general cultural factors. It’s worth noting that certain regions’ consumers are influenced to different degrees by marketing campaigns. Additionally, holidays vary by region, and you’ll need to decide whether to add this feature to the model or not. Take into account different legal constraints (limit for product amount) etc.

Perishable vs Non-Perishable Products

Finally, it’s crucial to factor product type within your demand modeling. With perishable products, you need to set up the right metrics to penalize the model when the prediction is much higher than the real value, as the consequences for excess inventory are significant. You need to be careful with data preparation and work with outlier detection because this prediction should be extremely sensitive to any changes.

Conclusion

The pandemic has made us all hyper-aware of the limitations and constraints of forecasting models, but the best practices in forecasting are the same whether we’re in a pandemic or not. You should always be striving to fit your particular business model and business needs to your forecasting tools, to attract new customers, increase your revenue, and expand your market share.

To find about more about practical applications of machine learning models, pick up a copy of Eric Wilson’s new book, Predictive Analytics For Business Forecasting & Planning. Written in easy-to-understand language, it breaks down how machine learning and predictive analytics can be applied in your organization to improve forecast accuracy and gain unprecedented insight. Get your copy. 

Below are the updated forecasts of coronavirus cases and deaths in the USA. The forecasts are daily rolling forecasts, looking 2 months ahead. Alongside the forecast, you will find actuals and forecast accuracy expressed as MAPE. This rolling forecast is updated weekly.

These forecasts are generated by Dr. Chaman L. Jain, Professor of Economics at St. John’s University, and author of the book, Fundamentals of Demand Planning & Forecasting. For further information on this project, including forecast assumptions, click here.

Download (DOCX, 20KB)

“Change is the only constant in life”, said the Greek philosopher, Heraclitus. Uncertainty is a derivative of change and has been a constant ingredient of the demand planning process. Thus, uncertainty in business is not new but the pace at which uncertainty is building is unprecedented.

Uncertainty in business is increasing due to the increase in options available to consumers in terms of new products and different channels, and occasionally worsens dramatically when events like trade regulation overhauls, economic recessions, social unrest, and pandemics like we are experiencing now occur. So, whether by choice or by necessity, our product and channel mixes can become disrupted presenting major challenges for Demand Planners.

Understanding Uncertainty In Your Product Mix

The good news is that there are ways to reinsert certainty back into business planning. First we must  comprehend the uncertainty we’re dealing with and draw a perimeter around it. This can be achieved in the following ways:

1. Segment Your Products

First, understand the scope of uncertainty and drive segmentation. The Covid-19 pandemic has been the biggest driver of uncertainty in recent times. However, it has not had a uniform impact on all products and services across organizations. On one side, where products like gym equipment and frozen foods have seen sky-rocketing demand, the hospitality and tourism sectors have drowned. There have been products and categories like fresh foods, home and personal care, consumer electronics that have stabilized after an initial knee-jerk reaction.

The first thing Demand Planners should perform is segmentation of products, services and customers based on the value they contribute, i.e. revenue, profitability, stability of the segments, and by value they seek e.g., quality, innovation, cost, and agility. Segmentation will allow planners to improve focus, reduce noise from the demand signals, and drive technical improvements to predictions and drive a collaborative response in unpredictable segments.

2. Understand the Assumptions Behind Your Demand Signals

Second, drive clarity on the assumptions associated with the demand signals. Demand plans and actual sales are both outcomes of multiple factors, assumptions, and decisions. In times of uncertainty, it becomes doubly important for planners to tag the changes in their demand plan with factors like internal and external events; product, customer and supply chain related assumptions; and, strategic and operational decisions made by the organization.

Clarity on these factors allow advanced technology to apply AI/ML algorithms to generate improved forecast accuracy and allow planners to identify the impact of various factors on the demand plan.

3. Assess the Agility Of The Supply Chain

And third, assess the agility of the supply chain. While the uncertainty on consumer demand directly and adversely impacts demand planning performance, the uncertainty on the supply side – plants, suppliers, warehouses, transportation services – directly impacts the responsiveness of supply chains to react to changing demand. Like the demand side, there are segments in supply chain as well, which operate at different levels of upside and downside adaptability.

Some of your plants, production lines, and suppliers may be suitable for a continuous manufacturing operation while others might be more flexible and accommodate different products with minimal changeover cost. It is essential for Demand Planners to understand these segments in the supply chain and shape the demand plan based both on constraints and dependencies on internal and external partners.

The aforementioned activities help us comprehend the uncertainty we’re facing and allow organizations to create a value-for-the-organization vs. value-for-the-customer matrix, which will sharpen our focus on the relevant segments.

Injecting Certainty back Into The Business

Following the assessment of uncertainty, the following 4 steps can be taken to drive certainty in the demand planning process and outcomes.

1. Update Demand Plans Regularly

As the first step, Demand Planners must establish a frequent cycle of plan refinement in the short-term. This can be a combination of demand sensing based on daily order positions or POS information, and consensus planning for key segments on on a weekly basis. The key to success is understanding of the internal and external factors driving sales and the ability to incorporate the impact of external factors in the demand plan.

The goal of short-term plan refinement is to get the deployment of products in the network right. As a by-product, it may improve forecast accuracy as well. Another important thing to note is that the periodic demand planning cycle should not lose sight of the big picture of achieving the annual operating plan.

2. Document Risks & Opportunities In The Demand Plans

The second step is that the demand planning process must improve the governance around recording of risks and opportunities and their probabilities. Risks and opportunities are the most ‘certain’ aspects of uncertainty. In current times, we are seeing customer and channel stability risks, product vitality risks, fulfillment lead time variability risks, and so on. On the other hand, we are seeing innovation-driven opportunities in product channels; the eCommerce channel is enabling leaders in the CPG industry to reach consumers directly, efficiently, and profitably.

With such risks and opportunities, Demand Planners must establish risk and opportunity governance and their inclusion in the normal, aggressive, and conservative demand plans in collaboration with customer facing roles. The organization must develop a response plan in collaboration with supply chain roles for each of the three IBP-generated demand plans to ensure agility when the business situation changes from one plan to another.

3. Foster Certainty By Doubling Down On Your Cash Cows

The demand planning team can drive certainty in the demand plan by playing to their strengths. Organizations can use a growth-share matrix to identify their Cash Cow and Star products and customers, and double down on them.

The supply planning team must ensure that the warehouses, plants, and suppliers that cater to demand for such products and customers are resilient, responsive and generate positive financial value for the organization. Together, the demand-supply and the value equation should drive the supply chain capacity decisions and subsequent constrained demand plan.

4. Engage with Suppliers

Finally, in the fourth step, the demand planning team and associated functions in the organization must collaborate with the broader supply ecosystem to gain clarity. The information coming from upstream and downstream ecosystem partners under the CPFR framework brings early insight into demand changes and supply risks across all nodes of the supply network. It provides additional lead time to collaborate and respond to uncertain situations.

In conclusion, while change may be the only constant in life, understanding the reasons driving change go a long way to driving certainty in demand planning.

While forecasting has always being associated with uncertainty, we are living during an unprecedented time that requires extra special handling. This article attempts to analyze the impact of Covid-19 beyond the near term, focusing on the entire year horizon.

As we are in the middle of the pandemic, it is difficult to know exactly how to react and what to prepare for, but there are certain things we can anticipate in terms of how consumer behavior will change over different time horizons, and how that will affect our forecasts and plans. Here I present a framework to understand how Covid-19 and lockdown impacts consumer behavior, and subsequently, the demand impact on your product portfolio.

Segment Products According To Necessity

During a crisis, product demand can shift from certain products to others as some items become more important to your customers or consumers than others. It is important to organize your products based on necessity and know how their demand is changing in this new environment. While there are a lot of interesting insights about consumer behavior during this current period of quarantine which is affecting demand, for the purposes of this article we will look at the longer-term impact of the current situation on demand. The below table below is organized to show anticipated impact on demand according to product need, both during and after the crisis.

Planning For High Necessity/Essential Products

High necessity goods can be categorized into three buckets based according to the demand during this pandemic.

Situational High Need Products: Considered as essential by consumers in order to “fight” the crisis which, in the present situation, are associated with flu-prevention and alleviation, including but not limited to masks, sanitizers, and medications. Excessive demand for such products has led to scarcity and re-shifting of supply to respond to this newly-formed need. Demand for such products will continue to be high until consumers feel “in control” of the situation and their demand is fulfilled because products are available for purchase and/or they no longer have to respond to the situation that caused the pandemic because a vaccine is available. If you have high situational need products in your portfolio, plan on constant demand sensing and longer-term gradual demand reduction.

Non-perishable High Need Products: These are considered as essential by consumers in normal circumstances and have a shelf-life long enough to last the isolation period. Examples include packaged and long-lasting canned goods, consumable paper products, and cleaning supplies. Excessive demand for such products has led to scarcity due to consumers stocking up, causing the so called “pantry loading” effect. Given how long these non-perishable products last, it is possible we’ll see a reduction in demand post-crisis, caused by consumers who stockpiled such goods at the onset of the pandemic. If you have non-perishable high need products in your portfolio, plan on the “pantry loading” effect wearing off consumer preferences shifting to other products.

Perishable High Need Products: These are considered as essential by consumers in normal circumstances, but unlike non-perishable/long lasting products, have an expiration date and/or special storage constraints. Examples are fresh food like fruit, vegetables and meat, which saw intermittently high demand as consumers were stocking up in anticipation of lockdown. Consumer can increase the “lifespan” of such goods through freezing but to a limited extent as there refrigerating storage constraints. If you have perishable high need products in your portfolio, demand to return to normal levels, but plan on a potential shift in consumer preferences.

Planning For Medium & Low Necessity Products

Medium necessity products are those not essential to sustaining life but considered by consumers to be essential to satisfying psychological needs. Some of these products saw the biggest level of demand transference due to store closures and companies limiting their operations. Medium necessity goods are categorized into two buckets:

Short-Lasting Consumption Medium Need Products: Examples include books, movies, games, office supplies, wellness and beauty products, and seasonal merchandise. As consumers were, and will continue to be, more isolated in their home, demand for such products will remain elevated until they feel safe to return to their usual activities.

Long-Lasting Consumption Medium Need Products: These products have an extended consumption lifecycle like sports gear, office equipment, and basic home entertainment and improvement supplies. As consumers were forced into a “do it yourself situation”, i.e. cutting their own hair, working out at home, carrying out car maintenance etc.,  they frontloaded this demand that which might lead to potential reduction until the effect wears off.

If you have any medium need products in your portfolio, plan on constant demand sensing and be prepared to respond to any demand shifts.

Low necessity products: These are considered as “extras” by consumers like high fashion, expensive merchandise not serving stay at home or outside safety functions. Demand for such products has decreased due to anticipated economic instability and decreased spending power in light of increased unemployment and will continue to be low until consumers gain confidence and purchasing power returns.

Post Crisis & Demand Shift & Planning Ahead

While demand at the time of disruption depends heavily on product necessity, some things remain the same. The shift from buying in physical stores to online is applicable to all products irrespective of their necessity. Triggered by government mandated store closures, stockouts in physical stores, and individual safety concerns, consumers adjusted their purchasing channels, which has compounded an existing trend, with many more people buying online. This will continue post-crisis as a significant portion of these consumers will stick to this channel, preferring the convenience of the online shopping over physical stores. We must anticipate a medium- and long-term demand shift towards the online channel.

As a result of stockouts, consumers were forced to buy branded and non-branded items that they wouldn’t have bought under normal circumstances. A portion of these consumers will not return to their pre-crisis brand preferences.

Similar to the way we would handle a dramatic events such as a hurricane (but on a much larger scale) demand planners need to ensure that excessive or depressed demand caused by Covid-19 is associated with distinct causal factors and/or be tagged as outliers, depending on the forecasting solution used by the organization.

• If your organization uses sophisticated causal based algorithms for demand forecasting, it is recommended to break down the Covid-19 period according to geographical location and supply the forecast engine with regional Covid-19 data (these are our dramatic impact attributes), allowing us to develop measures for each region according to outbreak severity.
• If your organization is relying on time series techniques, you will need to mark data points during Covid-19 as outliers. But this needs to be done post-crisis, allowing demand forecasts to react to recent trends for the duration of the pandemic.

While there wasn’t much we could do at the outset of the Covid-19 outbreak, there is a lot we can do as we exit the crisis and in the period following it. These periods require careful planning and consideration. As demand planners, it is in our DNA to deal with uncertainty and even if the current situation is a lot more extreme than we are used to, our usual approach of combining the “art and science” will allow for robust short, mid and longer term forecasts. Plan on continuous demand sensing based on product necessity and anticipate demand shifts for the year ahead and be aware that as demand returns to “normal”, there will be new consumer trends to plan for.

This article originally appeared in the Spring 2020 issue of the Journal of Business Forecasting. Click here to become an IBF member and get the journal delivered to your door quarterly, as well discounted access to IBF training events and conferences, members only workshops and tutorials, access to the entire IBF knowledge library, and more.

As we move into this next phase of this Covid-19 global pandemic with countries, states and businesses beginning to open back up, the importance of sound judgement and forecasting is critical. With these new times will come new factors, new information, and potentially new disruptions.

At the same time, we can learn from what we’ve been through so far as we begin to plan for today and what will come post crisis.

Based on our town hall meetings, talking with experts in the field from around the world, best practices, and my own experience, I have extracted these 7 lessons for navigating forecasting and planning as we come out of the pandemic.

1. Focus On Data

The past few months may not be representative of what is to come but we shouldn’t throw out that data. What we need to do is better understand the multiple drivers and begin to cleanse some of this data. We need to see what we can learn from it and possibly even discover new external variables that drive our business forecasts. Now, as we move into this next phase, is a critical time to evaluate the data we have and continue to collect as much data as we can that will help us in the future.

There is also bias and emotion that you need to filter out when faced with new, untested information.

2. Don’t Overreact

This has a couple of meanings for the Demand Planner. First, be careful with statistical black boxes or best pick forecast models. Many of these overreact to recent history or new information without enough consistency and skew trends up or down. Secondly, consider yourself and your own reactions. As we begin this next phase of this crisis there are still many unknowns. There is also bias and emotion that you need to filter out when faced with new, untested information. It is important to try to understand what information is going in and not overreact to what’s coming out, even if it’s negative.

As Erin Marchant, Associate Director for Global Demand Planning at Collins Aerospace, said in a recent article “Don’t panic…The organization will need level heads to lead in a time of uncertainty, and demand planning can provide this leadership.”

Latency in forecasting can be costly and we need to be able to update plans often and efficiently.

3. Update Often

We have been witnessing business and life change at an astounding speed. There is no sign this will change anytime soon. As things change daily, you need to develop a rapid reporting capability or, if you have such a capability in place, keep reporting. Even though we want to make sure we do not overreact to new information, we still need to capture and understand new information as it is made available. As we continue to navigate new consumer demand and environmental factors, ad hoc planning and the running of multiple scenarios will be needed.  Latency in forecasting can be costly and we need to be able to update plans often and efficiently.

Demand is not going to start looking like your historical data anytime soon so you need to look at the why consumers are making those purchases.

4. Continue To Ask Why (Predictive Analytics)

One of the big lessons companies are learning during this pandemic is that, in addition to forecasting with historical sales data, we need to look at external variables as well. Demand is not going to start looking like your historical data anytime soon so you need to continue to look at the why consumers are making those purchases, and why sales for certain products have risen while others have fallen. This is predictive analytics and maybe this pandemic is the impetus you need to get started with something you should have already been doing, or at least thinking about. These answers usher in a new era of predictive analytics and a forward-looking perspective that will help us navigate through this.

5. Don’t Focus Too Much On The “What”

The uncertainty has been horrible and it’s not getting any better. This correlates to poor forecast accuracy and sometimes bias. Measuring uncertainty by measuring forecast accuracy is important to tell other functions what to plan for and risk but does little to tell you how good of a job you are doing – right now your forecast accuracy will be poor and there isn’t too much you can do it about. Don’t focus too much on these numbers, rather focus on how you can use those numbers to improve your processes.

6. Communicate And Be Transparent

One of the best things a Demand Planner can have during these times is humility. We will be wrong, but what we are providing is a best estimate of what is going to happen based on data and collaboration. Do not be afraid to explain the uncertainty and do not shy away from forecast performance. Communicate a data-driven assessment and be transparent when it comes to risk.

7. Prepare For Possible New Norms

Part of humility is the willingness to change. We should expect that because of Covid-19 things will change. Consumers may purchase differently through different channels. Companies you deal with may change inventory policies or buying habits. The forecasting models you relied on no longer work. We must understand and not only prepare for possible new norms but be willing to abandon old assumptions that no longer provide the same results.

8. Get Ready For The Next Black Swan

While we cannot truly predict a black swan, we can see its shadows. Not surprising, history does, and will, repeat itself again and again. Rather than letting out a big sigh of relief and returning to normal routines when the crisis subsides, efforts should be made not to squander a valuable learning opportunity. Collect data now, focus on external variables and more predictive analytics now, and focus on a solid demand planning and forecasting function and best practices today. The one thing we have learned if nothing else is that preparing for the next crisis (or the next phase of the current crisis) now is likely to be much more effective than an ad hoc, reactive response when the next crisis hits.

The following is a daily rolling forecast of Covid-19 cases and deaths in the USA and New York State, looking 2 months ahead. It is prepared by Dr. Chaman L. Jain, Professor of Economics at St. John’s University, and author of the book, Fundamentals of Demand Planning & Forecasting. This forecast will be updated weekly as new data emerges. 

When preparing a forecast for something new, whether it’s a product or a virus, we typically identify an analogous “item”. We identify how the analogous item behaved in the past to predict how the new item will behave in the future. But the patterns of Covid-19 cases and deaths do not correspond with any virus we have experienced before, making this impossible. Further, the patterns in countries like South Korea and China that have nearly gone through the whole coronavirus cycle, do not match with what we are currently experiencing in the U.S.A. Therefore, the only option we have is to study the pattern of cases in the U.S.A. and then extrapolate it going forward. We now have enough data to do so.

Forecasting Coronavirus Cases & Mortality In The United States

Usually, the pattern of a virus (like a new product in business) forms a S curve where it first increases at an accelerated rate and then increases at a decelerating rate. This is exactly what we are seeing in the U.S.A. Covid-19 data. Data shows that we reached the point of inflection in the week of March 16^th, a key turning point when the daily percentage increase in total cases started increasing but at a decreasing rate. In that week, the weekly average of daily increases as a percentage of total cases hit 38%. Thereafter, it started declining and fell to 3.6% in the week of April 20th. I believe this pattern will continue and that the total number of cases in the U.S.A. will hit 1.7 million by June 30^th. After that, we will still have cases of coronavirus, though their number will be much smaller.

The U.S death rate from coronavirus also follows a similar pattern. It reached its point of inflection in the week of April 20^th when the weekly average of daily deaths as a percentage of total cases reached 0.2%. I expect this percentage will continue to slowly decline. With that, I expect the death toll in the U.S to reach 170,000 by June 30^th.

New York State Forecasts

Among all the states, New York state has been hit the hardest. In this state, the pattern of people affected by the virus is very similar to that of the  U.S.A. as a whole. The weekly average of daily percentage increases in total cases peaked in the week of March 16^th when it rose to 58%. Thereafter, it started declining and reached 2.5% in the week of April 20^th. I expect this pattern to continue and that the number of cases in New York will reach 385,000 by June 30^th.

Regarding the number of deaths in New York state, the pattern is the same as the total number of cases. The daily number of deaths as a percentage of total cases kept on rising until the week of March 30^th. Thereafter it declined and is expected to decline further. With that, the total number of deaths in New York State is predicted to reach 30,000 by June 30th.

It should be noted that every forecast is based on certain assumptions. A key assumption here is that things will continue the way they have in the past. During the time period observed to create our forecasts, no vaccine to treat this virus was available. The development of such a vaccine would cause us to revisit our forecasts.

 Daily forecasts are provided in Table 1. One can observe how accurate they are by comparing each day’s forecasts with actuals.

 Check back next week for the updated forecast.

SPECIAL REPORT: Coronavirus disruption to sales forecasting has made an already complex process seemingly impossible. Until a vaccine is widely available to the public and infection rates are under control globally, businesses face many confusing demand signals from their own data and a barrage of ever-changing news updates. Navigating supply chain disruption amidst a pandemic is confusing everybody, including the experts.

This article offers guidance on S&OP process management to help your teams make clearer business forecasting decisions during the current Coronavirus pandemic. For small businesses and international corporations alike, this article offers examples and demonstrates the value of integrating new data and gaining new insights for optimal strategic and operational decisions to survive the Covid-19 pandemic.

Sophisticated tracking devices to monitor different aspects of your operations offer real time data, illustrating moment to moment changes. Shortcomings in supply chains can be explored, and scenarios split tested against each other for comparison for flexible, speedy decisions. Over the course of the pandemic, tracking evolving consumer behavior and changing supply chain capabilities help build the new baseline forecasting assumptions we need.

Incorporating Changing Assumptions Is A Must

Patrick Bower, Senior Director, Global Supply Chain Planning and Customer Service at a multinational consumer goods company, believes that scenario planning depends upon the specific needs of each business. He uses tools to help balance emerging changes in supply and demand. For example, if a 20 percent increase in demand occurs, capacity impacts are investigated immediately and resources are planned accordingly.

For some businesses, capacity in terms of available personnel as well as supply chain disruptions inject unanticipated consequences into planning, at least until the pandemic is under control. While tracking a fall or lift in demand or supply will already be a familiar continual assessment for many, forecasting now demands integration of government policy on social distancing and availability of a vaccine into analyses. Covid-19 related events represent new indicators to build into time-series forecasting. Incorporating new assumptions is critical and simply looking at the past no longer works. 

The point is to “get comfortable” in knowing you may be wrong

The point, says Bower, is to “get comfortable” in knowing you may be wrong in scenario planning under the present circumstances. Nevertheless, he acknowledges that being data focused offers the best insurance against error. Interpretations of recent data may suggest “what if” questions and testing scenarios, highlighting the gaps that a collaborative S&OP process can fill. 

As there is a great degree of uncertainty given a lack of data and rapidly evolving events, he suggests collaboration with external stakeholders wherever possible to gather (and share) as much information as possible. Supply chain partners need to be supported. An example is sharing POS data with them which helps improve their planning which, in turn, helps secure the supply you need. For consumer goods companies, there is value in contracting market research partners to guide your risk management. Insight into consumer behavior at a time like this is King.  

His team reviews potential “weak links” in supply chain data projections. During a pandemic, where government policy surrounding lock-down is unclear, some companies may not define themselves as an “essential business”. Suppliers that have identified themselves as non-essential businesses and have shut down are a serious problem for many companies. Depending upon the Covid-19 trajectory, more “weak links” like this in the supply chain could unfold.

Collaboration and Communication Key As Judgement Comes To The Fore

Collating and interpreting novel internal data, flagged by colleagues, particularly those on the front line, as well as supply chain partners, could be essential to enhanced and agile decision making. Crises offer opportunities for staff contributions to identify new performance markers and future indicators during disruption. For Andrew Schneider (ACPF), Manager of Corporate Quality at Medtronic, transparency is key, and new internal and external relationships must quickly be forged to ensure timely production and delivery of products.

Where machine learning does not suggest appropriate substitutes, companies have to use their best judgement, unless alternative suppliers can be mobilized rapidly.

Demand planning software systems must facilitate integration of up to date information from upstream, where products may be drying up, as customers switch lines. Where machine learning does not suggest appropriate substitutes, companies have to use their best judgement, unless alternative suppliers can be mobilized rapidly. Weighing risk and acting accordingly should involve continual monitoring of implications of changes made.

Regular Monitoring & Tracking Of  Data Is Critical

Any tweaks to procedure need to be systematized for close monitoring within key S&OP cycles, which vary between businesses. Small adaptations can be tested against emerging data to review impacts. This necessarily involves open communication with relevant stakeholders for the benefit of all moving forward, including end users.

Holding onto life-saving products is not only immoral but can damage business reputation.

Dramatic operational changes may also be entirely appropriate. However businesses choose to adapt, close observation and consistent, regular tracking of results is essential. Comparisons against data from economists and epidemiologists as well as against data from previous disruptions are recommended. Cross-functional teams need to support interpretation of forecasting results, facilitating rapid decision making.

For retailers panicking about lack of inventory, it is also worth bearing in mind that it may be entirely justified to run out of stock, such as face masks, or bleach. During this catastrophe, say Patrick Bower, holding onto life-saving products is not only immoral but can damage a business’s reputation.

Now’s The Time To Use Wider Indicators

Individual companies will have individual balancing acts and assumptions to include in their forecasts, focusing on a wider variety of key indicators than usual. If cash flow, as opposed to inventory or service levels, is the main priority, then demand planning managers will benefit from integrating wider indicators, such as the shape of a forthcoming recession/recovery, for instance. Segmenting historical data sets according to test scenarios around a ‘V’, ‘U’ or ‘W’ shaped recovery will reveal implications for S&OP and cash-flow. 

However, given that time series forecasting cannot predict unprecedented events, disruptions like staff absenteeism, supplier or line loss, and even switching to producing a new product, requires using cleansed historical data. Data can be split tested in forecasts allowing implications to be explored before decision are made.

Jonathan Schwartz (CPF) is a Supply Chain Analysis Manager at WD-40. He remarks that the baseline ‘steady state’ looks different depending upon a company’s fiscal year – forecasts for April-end could look good, but not so if your year end is December. He adds that fast production and distribution is essential before absenteeism from sickness or changes in business partner behavior disrupts either business function.

While we wait for a vaccine, confidence and behavior will continue to shift, changing consumer, supply chain and staff priorities.

While we wait for a vaccine, confidence and behavior will continue to shift, changing consumer, supply chain and staff priorities. This requires daily, weekly and monthly reviews of demand variables, KPIs, macro-economic indicators, and the spread of Covid19.

Matt Hoffman at John Galt Solutions believes 12 month planning to be a key timeframe as companies must be must be positioned appropriately when things return to normal. During these initial stages in the pandemic, where social distancing is the norm, there will be pent up demand. As businesses ‘return to business as usual’ environments, regular re-assessments of assumptions will be necessary before forward planning. It is recommended that companies understand in detail their inventory carrying plan for this next year (during which time there may yet be a second wave in the pandemic) as lock-down restrictions are lifted.

Make no mistake, Coronavirus has changed consumer behavior and some of those changes are here to stay.

When combining data sets in scenario planning, John Galt Solutions observe income and consumer confidence, deploying regression modelling for understanding consumer impacts during these times of social change. He cites health and beauty product consumption shifting from salons to home application under lock down. Thus price points and or marketing messages need recalibrating. Make no mistake, Coronavirus has changed consumer behavior and some of those changes are here to stay.

Forecasts Will Be Wrong & That’s OK

Industries and businesses are at risk during the current unprecedented circumstances. However Coronavirus and the responding policies develop, and whatever the impact on the economy, experts are consistent in their message: Closely monitor the data and compare against historical data from previous disruptions and downturns. Furthermore, collaboration and communication in demand planning have also never been more necessary. If S&OP as a collaborative, cross-functional forum was important before this crisis, it is a life saver now. 

Forecasting models will not be “correct” in the near term.

During a potentially dangerous new phase as world leaders to seek to balance public safety with a return to work, the coming weeks will provide yet more tests of companies’ forecasting and planning abilities. As Eric Wilson, Director of Thought Leadership at the Institute of Business Forecasting, notes, the concern of many of the businesses contacting him is the duration of disruption. This shines a spotlight on the importance of looking beyond sales data and integrating economic and epidemiological data into forecasting.

He adds that while forecasting models will not be “correct” in the near term, it is times like these that reveal how critical forecasting and planning are to a company’s survival.

Useful Articles:

Demand Planning During A Recession

Planning During A black Swan Event

Supply Chain Planning During Covid-19

3 Veterans Give Advice On How To Plan For Coronavirus

The Impact Of Coronavirus On Your Forecasts