Q. What is the best method to forecast in the toy industry ? Or any industry where 70% + of the portfolio is renewed every year across thousand of SKUs ? Aggregation is understood, problems starts when entering SKU territory .

A. There is no special model for the toy industry. There are three types of models: (1) Time Series, (2) Cause-and effect, and (3) Judgmental models. They apply to all industries. We use the model that best captures the data pattern. SKUs are always most difficult to forecast. Usually, we first make a forecast of the category, and then allocate the share of each SKU using the ratio of each SKU to the total based on the rolling average of the last 12 months or so. That is, what the ratio of SKU 1 is of the total category, what the ratio of SKU 2 is of the total category, and so on. By using these ratios, we can make a forecast of each SKU.

I hope this helps.

Happy forecasting!

Dr. Chaman L. Jain

Editor-in-Chief,

Journal of Business Forecasting

In the last four decades or so, more and more companies have gone global. With that, the structure of the S&OP process has changed, though the key objective remains the same—aligning supply with demand, integrating operational plans with strategic plans, and optimizing product portfolio and product mix.

What type of Global S&OP structure we need depends on the company’s supply and demand configurations. There is no one-size-fits all. Some companies have standardized products where the same products are sold all over the world; others have region specific products with different requirements for labelling and packaging. Some companies have manufacturing plants located all over the world, while others have them only in their home country.

So, different configurations of supply and demand require a different structure. Here we discuss a generalized Global S&OP process, which can be adapted, with some tweaks, to different situations. The process is typically done on a monthly cycle.

The Global S&OP Process

 The Global S&OP process discussed here assumes products are produced and sold globally. The regions involved are: (1) North and South America, (2) Europe, (3) Middle East and Africa, and (4) Asia-Pacific. It is a 6-step process, which works as follows (see also Figure 9.1):

Step 1: Regional Demand Review

Each country prepares and reviews demand forecasts with a team comprised of functions such as Sales, Marketing and Finance, and sends them to the Regional Demand Planner. The Regional Demand Planner consolidates all the demand forecasts of its region, meets with its team consisting of the Regional Marketing Manager, Regional Sales Manager, and anyone else involved.

If necessary, he or she makes adjustments in consultation with the team and then sends the forecasts of each category to the Category General Manager (GM). (This is needed only where a company has a number of large categories. Otherwise, the Regional Demand Planner will send the forecasts directly to the Global S&OP Demand Planner.) Regional forecasts are better because they are closest to the market, and local teams know better than anyone else about their customers, competition, and consumers.

Step 2: GM Category Review

The GM of each category reviews demand forecasts and, if needed, makes adjustments in consultation with the Regional Demand Planners. The GM then consolidates each category into a global forecast and sends them to the Global Demand Planner.

Step 3: Global Demand Review

Here the Global S&OP Demand Planner reviews forecasts of each category with its GM, makes adjustments if necessary, and sends them to the Global Supply Planner. (Where Regional Demand Planners send their forecasts directly for Global Demand Review, the Global S&OP Demand Planner will make any necessary adjustments in consultation with the Regional Demand Planners.)

Step 4: Global Supply Review

Here the Global Supply Planner reviews the demand forecasts to see if they can be met. If there are gaps, they discuss how they can be closed. Based on the demand forecasts, the Global Master Production Scheduler develops a production plan, considering outstanding inventory. For that, the scheduler uses an optimizer tool which is programmed to maximize gross profits based on total delivered cost to customers.

The plan includes how much is to be produced and at which plant. The Global Supply Planner also prepares a report on misses and their possible causes. Misses include where we are likely to go over and below the demand. All this information is then sent to the Global Pre-Executive team.

Step 5: Global Pre-Executive S&OP Meeting

Participants of this meeting are the Global Demand Planner, Global Supply Planner, and Category GMs. The objective of the team is to come up with a plan that helps to align supply and demand, meets financial targets, optimizes product portfolio, hits market share targets, and/or fulfills any other goal that is important to leadership.

Wherever there are gaps, the team tries to close them. Although the team tries to resolve most issues, those that remain unresolved or for which consensus could not be achieved, are placed, among other things, on the agenda for the Global Executive S&OP meeting.

Step 6: Global Executive S&OP Meeting

This meeting is attended by key people including the President, high-level executives, Global Demand and Supply Planners and the CFO. They evaluate the plan presented by the Global Pre-Executive S&OP team against the company’s strategic plan, policy, and risk parameters. It also makes sure that maximum profitability within service and inventory constraints is maintained. If needed, adjustments are made to the plan. Once approved, it is implemented. If there is a significant change, the Global Master Production Scheduler is asked to re-do the production plan.

Where the company has different business units and each one is large, it may decide to have a separate S&OP process for each. The Global Master Production Scheduler, based on consolidated demand forecasts for all business units, allocates production among different units in an optimal fashion. The production allocation includes what to produce at which facility, and how much.

In Summary

S&OP is a great tool for managing demand but in a globalized world, emphasis is now on global maximization of financial performance rather than local or regional. For global maximization, we need a global process. Without it, different regions and countries will be more inclined to optimize performance of their region or country, and not of the company as a whole.

Further, a global process provides increased visibility across the entire supply chain, facilitating better and quicker decisions. It also improves communication and teamwork across regions, making it easier to manage tradeoffs.

The above is an excerpt from Dr. Chaman Jain’s Fundamentals of Demand Planning & Forecasting, widely recognized as the most comprehensive book written in the area of demand planning and forecasting. Get your copy here.

S&OP, the topic of this special issue, is a great process for managing demand, which also goes by the name of SIOP (Sales, Inventory and Operations Planning), IBP (Integrated Business Planning) and MIOE (Merchandizing, Inventory and Operations Execution. It doesn’t matter what we call it; the priority is to have a framework that facilitates cross-functional collaboration that serves to balance demand and supply while meeting the strategic goals of the business.

I needn’t mention the impact of COVID on our planning which has forced us to shorten the S&OP cycle from monthly to weekly. Make no mistake, S&OP saved many companies from the brink of disaster since the pandemic struck, as supply shortages and shifting demand wreaked havoc on our businesses. I was pleased to see that planning professionals made adjustments to their usual process to react to pandemic-related disruptions because there are longer term factors that are testing S&OP’s ability to manage demand, which also require new ways of thinking.

S&OP is tried and tested but it is more than 30-years old. Since it was first developed, market dynamics have changed, requiring changes in how we plan. Markets are now demand, not supply driven. Competition is intense. New products are exploding, and so are the channels of distribution. All these have added uncertainty that must be addressed. COVID or no COVID, we must ask ourselves how we can evolve S&OP to response to these shifts.

The increasing importance of new products requires a change to the traditional S&OP process. Although new products are reviewed within the product portfolio, they don’t get the attention they need. A significant amount of revenue comes from them and is growing. McCormick USIG, for example, gets 35% to 45% of its revenue from new products; LEGO, 60%; and Hasbro, 80%. To manage their demand, they require not only more attention, but also special skill sets. The success of new products depends on a robust projection of future sales, but they are difficult to forecast because of lack of history. One way to do it is to prepare three sets of forecasts: frozen (where no change can be made), slushy (where a limited amount of change can be made), and liquid (where any amount of change can be made). The other way is to prepare high and low forecasts — low forecasts for fixed contracts and high for flexible contracts. An S&OP step dedicated solely to new products is required.

Another component that needs to be added to S&OP is eCommerce, which is rapidly growing. If S&OP was conceived around brick-and-mortar sales, eCommerce requires a different skill set and strategy. For example, in eCommerce, customers buy less but more frequently. How they respond to a 24-hour online-flash sale is very different from brick-and-mortar. Since Demand Planners have access to customer data (e.g., how much they bought and when), they can develop a better marketing plan based on recency, frequency, and monetary value. Further, it is much easier to do demand shaping in eCommerce. eCommerce requires more agility in the supply chain because orders must be shipped on time, otherwise, it will lose the sale.

Further, in the omnichannel environment, there are a number of ways a product can be bought and shipped: (1) buy from a store; (2) buy online, ship from a distribution center; (3) buy online, pick up at a store; and (4) buy online, ship from a store. Each option has a different effect on the bottom line because of differences in their operating cost — picking, packing, inventorying, and shipping. These are factors that must be clearly identified and planned for in a dedicated eCommerce step in S&OP.

To keep S&OP robust, we must recognize these issues and have a mechanism to deal with them. Then, perhaps, we can bring S&OP up to date and fit to deal with the advancements we have seen in recent years. You’ll find in this special issue a range of excellent articles designed to help you implement S&OP or to improve an existing process, written by some of the leading figures in the field. I trust you will find them valuable.

I suggest using IBF’s S&OP maturity model in conjunction with reading these articles. Available at www.ibf.org/sop-maturity-model, it’s a free self-assessment that identifies your current S&OP maturity level and provides recommendations to progress to the next level, as well as other helpful resources.

Happy forecasting!

Chaman L. Jain, Editor

Institute of Business Forecasting

cjain@ibf.org

This is an extract from the 2021/2022 Special Issue of The Journal of Business Forecasting on implementing and sustaining S&OP. Download an extended preview here or become an IBF member and get both full digital access and the print version delivered to your door every quarter, plus a a host of other member benefits including discounted conferences and training, exclusive workshops, and access to the entire IBF knowledge library. 

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

Q) What is forecast bias? Do you have a view on what should be considered as “best-in-class” bias?

A) It simply measures the tendency to over-or under-forecast. It is an average of non-absolute values of forecast errors. If it is negative, a company tends to over-forecast; if positive, it tends to under-forecast. In the machine learning context, bias is how a forecast deviates from actuals. In new product forecasting, companies tend to over-forecast. But for mature products, I am not sure. So, I cannot give you best-in-class bias. To me, it is very important to know what your bias is and which way it leans, though very few companies calculate it—just 4.3% according to the latest IBF survey. If we know whether we over-or under-forecast, we can do something about it.

I hope this helps.

Happy forecasting!

Dr. Chaman L. Jain

Editor-in-Chief,

Journal of Business Forecasting

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)

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.

Q: Thanks for the Wonderful Book “Fundamentals of Demand Planning & Forecasting”. Is there any best proven and Industry accepted model/algorithm to calculate sales loss? For example, the actual sale duration for an item is from 7 am to 10.30 pm every day. But if an item is sold out at 6 pm, how do we calculate the number of sales that we lost/missed during the remaining time period from 6 pm to 10.30 pm?

I am working as a Demand Planning Manager in the Food & Beverages and Health & Supplement industries. We do demand planning on a daily basis with 3 days in advance (T-3) for Sale day (T) {Freeze Demand plan on Monday for Thursday’s sales}. We need a minimum of 2 days (T-2 & T-1) for procurement, transport, cooking and other operations. Buffer stock/Safety stock is not possible as shelf life of cooked food is max 24 hours. So on any day, our forecasted demand is equal to sum of sales plus wastage.

Often we face difficulties in calculating the correct sales loss percentage due to item stock outs.

Below is the data (imaginary numbers):

Date Stockout @ Demand Sold Wastage Sales Loss
Day 1 No Stockout 100 95 5 –
Day 2 No Stockout 105 98 7 –
Day 3 No Stockout 105 100 5 –
Day 4 21:42:29 95 95 – ?
Day 5 21:34:40 100 100 – ?
Day 6 20:21:47 98 98 – ?
Day 7 20:09:54 100 100 – ?
Day 8 No Stockout 115 105 10 –

Is there any algorithm/method to calculate sales loss for the days 4 to 7 in order to identify the overall potential sales/day?

MAPE = {(Wastage+Sales Loss)/Actual Sales}*100

Without correct sales loss numbers, it is very difficult to get MAPE correct.

Thanks in advance.

A: Your problem is very unique, and thus you won’t find any textbook that discusses it. If I were to face this problem, I would take the average sales on days when there was no stockout. The difference between the average demand and actual sales is your loss. For example on day 4 sales would have been 106 but you sold only 95, so lost sales equals 11 units.

I hope this helps.

Happy forecasting.

Dr. Chaman Jain

St. John’s University

[ Q ] Do you have any research/survey data regarding where to place the forecasting function? I have a client who is in the process of migrating to the S&OP process and wants to decide where to put it. They do not want it within Supply Chain, and are looking for a data-supported alternative. Anything you can provide would be greatly appreciated.

[ A ] Based on an IBF survey, most companies house their forecasting function within Supply Chain (49%), followed by Sales. The reason may very well be that forecasts are used most by the Supply Chain. Further, data shows that forecasting error is the lowest when they have it within Marketing (16.30%), and highest in Supply Chain (26.32%).  You read more on the relationship between different departments and forecast error in IBF’s research report, ‘The Impact of People and Processes on Forecast Error in S&OP’.

[Ed: You can find all of IBF’s research reports here.]

Happy forecasting,

Dr. Chaman Jain,

St. John’s University

Question

Dear sir,

I’m working in a procurement department. I have an interest in learning demand planning, an area in which I have some doubts.

1 – What is consistent upward and downward bias. Where do we use these?
2 – What is constrained and unconstrained demand data? What are the benefits of each?

 Answer:

1 – Upward or downward bias is caused by the optimistic or pessimistic attitude of a forecaster. An optimistic attitude causes an upward bias by using optimistic assumptions in building a model which may be, for example, the economy is expected to grow in the next period at a healthy rate, a competitor is unlikely to respond to our promotional efforts, distribution and velocity of a product are expected to increase, and so on. And/or he or she has tendency to intuitively raise the forecast numbers. A pessimistic attitude does just the opposite. In new product forecasting, there is a tendency towards over-forecasting because the sponsors of forecasting are generally overly enthusiastic about the product.

2 – In constrained demand, demand is adjusted by applying constraints such as production capacity and material shortage. If you receive an order of 100 units, but, for one reason or other, you can only deliver 70 units, then the constrained demand is 70 units. The unconstrained demand is 100 units—the number of orders you receive irrespective of whether or not they can be fulfilled.

Happy forecasting,

Dr. Chaman Jain,

St. John’s University