The following is the second in a two-part series covering the art and science of  demand planning. Read the first part on the art of planning here.

Why do we Need to Forecast?

Companies need to understand what the future will bring to make the right decisions: What is my optimal footprint? What is the right price based on future expected volume? How much capacity is needed, and which inventory investments will yield the best returns? Using a statistical forecast provides a good starting point to answer all these questions.

However, not everyone has the background to understand the math behind the process, the underlying assumptions used in the forecast, or the best way to integrate this information into demand planning decision making. This is why it is important to simplify the statistical forecasting principles and assumptions at a high level so everyone actively participate in the discussion and help build the demand plan.

The Science Behind Statistical Forecasting

“The only way to predict the future is to understand the present.” – Isaac Asimov

The first question that needs to be answered is: What are you trying to predict? Forecasting order entry or sales will provide different outcomes and maybe both are needed, but for different reasons. My experience is that to drive the supply chain process you are better off using order entry as an input, since sales includes variability related to supply availability. But if there is a conscious thought process behind the decision of what is being used, it should not be a problem. Let’s review a quick example to clarify this situation.

A few years ago, during a demand discussion among colleagues, we were seeing a high level of volatility in a certain SKU. The Demand Planner was not able to explain it well as the item had steady sales in the past. Then the Supply Planning Manager stepped in and mentioned that we had a recent shortage in this product and the peak in sales should be due to a big lot coming in and being shipped out in the same month.

These circumstances lead us to review the process and the data we were using, as we were trying to determine what we could sell if there was not any supply constraint. The best set of data to do this was orders entered with the requested date from the customer. Once we used these figures, the pattern was much smoother as order entry had the information of what the customer needed, and the sales information had the information of when we were able to provide it. These subtle differences in the input of the process provide a very different output.

Now we can move to discuss the process. The intention of this section is to understand from a ten thousand foot view the components of the models and the high-level assumptions behind them. There will be complex calculations involved and we will not get into the details behind the math. After all, the experts can handle this very easily, but it is the understanding of these concepts in general that helps us to drive constructive discussion and decision-making.

There are a lot of different methods that try to predict the future including a simple moving average, exponential smoothing, econometric models, linear programming methods or machine learning algorithms. We will focus on time series – a series of sequential data points ordered by time. The two main methods we will review are exponential smoothing and linear regression because they are the easiest models to understand and the most widely used.

Exponential Smoothing

The first technique is called exponential smoothing. This method uses historical information to decide how much weight to give more recent history, trend or seasonality. These are the three main factors that you want to spend some time reviewing as they will give you some insight into the current demand pattern. The math will tell you how important each of these is, but it is up to people that understand the market to explain why.

This type of modeling is very helpful as it is data driven. I remember a meeting where a person on the marketing team mentioned that a certain product family was seasonal and after reviewing the data, it turned out that this was not statistically true. After digging into the details, it turned out that some products in that category had a strong seasonality component, but at the aggregate level you could not distinguish that pattern.

Another way to predict the future is to tag demand to macroeconomic variables and see if there is some correlation between them. A good indicator used in the plumbing industry is housing starts since it is a leading indicator of economic activity that tells you how many more homes are being built. There are a lot of different economic indicators out there and probably one of them is a good fit for your industry. Finding this correlation would be very useful since there are public forecasts available that can be used to predict the demand of your products.

Underlying Assumptions & Caveats

“History never repeats itself, but it does often rhyme” – Mark Twain.

What happens in an environment where there is volatility, uncertainty and complexity? Well, some time is spent discussing the data but this is usually not enough to understand all the caveats and assumptions.

One of the most important principles in forecasting is the underlying assumption that history will repeat itself and that past information can provide a good representation of what will happen next. The math tries to find certain patterns hiding in the data, like determining if the recent past is more useful or if there is a trend or seasonality involved. Early on in my career, when forecasting a high-volume item, we got a very high forecast for the near future, which did not made sense. It turned out that we had a big promotion in the recent past that was skewing the numbers. Once we took this out, the forecast corrected itself. The lesson we learned was that it is very important to scrub history to find outliers in the data.

Another important assumption is that external factors in the environment will remain constant, allowing the forecast to be developed under similar circumstances every time. Variations in the industry outlook, regulatory changes or economic growth fluctuations provide a challenge to the models described above. An alternate way to incorporate this information into the demand plan is required in these cases. A recent example of this is that after years of economic growth, the economy is stagnating or even shrinking. While this is known and discussed in the news and at the watercooler, there is always a lag from when this starts to happen and when the forecasting model picks it up.

Finally, a few important factors to consider in the assumptions are the horizon that you are using to plan and the level of aggregation. Think about the weather for a moment – usually the forecast for tomorrow is very accurate but looking at any day next month is not worth it. The same is true for any type of industry forecast; the further your look out, the less accurate it becomes.

It is a similar situation for the level of aggregation. It might be easy to predict how much you will sell in dollars for next month but it is harder to determine how many dollars per SKU or per customer you will sell next month. This is important to understand, because I have faced situations where the business asks how much we will sell in five years at SKU level by customer. It does not make sense to do a detailed analysis in this situation.  It is better to provide a directional number that is easy to explain and that supports the business to make the right decisions.

Measuring The Output Of The Process

Some of the most popular phrases in demand planning are “If I could predict the future, I would go to Vegas” or “The forecast is always wrong” and, my favorite one, “My crystal ball is broken”.  It is a fact that no forecast will be exact, but you can get within a decent range and with a good level of confidence. This is why it is good practice to measure the accuracy of your statistical forecast to understand the reasons behind your top misses.

From my perspective, it is important to understand how good the forecast accuracy is in terms of mix and volume. A good way of measuring mix is through MAPE (mean absolute percentage error), which basically tells you by how much you missed the forecast – regardless if you missed up or down – compared to what your actuals were. The advantage of this metric is that it is easy to understand (since it is a percentage) and it does not net out negative and positive values as it uses an absolute error.

The way to measure volume is through understanding if there is a bias at the aggregated level of your forecast. For example, when aggregating all your forecasts at the total dollar level by month, if you find out that the actual sales number has been below the forecast for some time, you might want to understand the reasons behind this. Ideally you want to oscillate between being a little above and then the following period a little below over the time horizon.

Discussing The Forecast During The Demand Planning Meeting

After understanding how a computer (or a Demand Planner) creates a forecast, it should not be a surprise why this needs to be reviewed in a group setting. Even if math is not your strength, a lot of value will come from having a thorough discussion of the historical data, the process used to come up with the numbers, the actual forecast, and metrics.

Below is a checklist of things to review. During this discussion, make sure that you balance time between the items that will move the needle, but covering in enough detail that allows you to understand the current situation.

• Start with the input. Are we using the right dataset and have we looked at history to review and scrub?
• How much importance do we give to more recent history versus the past? Is there a pattern in the data to be discerned, like trend or seasonality?
• Is there an economic indicator that we could peg to a group or family of SKUs that will help us determine what the future holds?
• Is there any external factor that could deter us from using our forecasting methods to determine the future values? Examples of this are changes in price (by us or the competition), changes in the economic or regulatory environment, or even new products that could cannibalize current demand.
• What are the forecast accuracy metrics telling us? Are we better than last month? What are the top offenders and why?

In conclusion, a good statistical forecast provides a good start for your demand planning process and a solid foundation. The data, assumptions and metrics need to be understood and discussed in depth. But keep in mind that this only the beginning of the process. There is an art component of the demand planning process that incorporates changes in the environment, market intelligence and in general a consensus between areas that should be aligned with the company strategy.

Amid the hype of data science, machine learning and AI, one important question arises within the supply chain world: Will any machine or algorithm be able to do the demand planning process by itself? Do we still need a full-blown process that requires people across different functions to interact with each other to create the best demand plan? The answer is yes. In the first of a two-part series that examines the art of science of demand planning, here we dive into the all-important art element that considers people, process and market behavior.

Why Do We Need Anything Beyond A Statistical Forecast?

While the science aspect of demand planning is very important, there are some key things that an algorithm cannot do yet. It cannot understand competitor or customer behavior, or determine the impact of external economic dynamics that could potentially affect the business. Nor is it capable of factoring in uncertainty and risk into the decision-making process.

A good forecast algorithm will eventually catch up with some of the situations described above, as customer orders or sales being used as an input to the process are affected by them. Calculations are usually rigged to put more weight on recent history as soon as a trend is detected. But there is a lag before this happens and sometimes waiting until this is noticed by a statistical model can be very costly for the business, since every month that passes compounds the issue, since the effect is cumulative. If this happens any remedial action has to be much more drastic than if smaller adjustments were made over time.

This means that the past is not always a good predictor of the future, and that human intervention is needed to overwrite the statistical forecasts and make judgement calls that avoid major remedial action down the road. The best way to execute this is through an effective alignment across functions that allows the company to stand behind a single plan that everyone believes in. This, in a nutshell, is the art component of demand planning.

Understanding the Market and your Customer

In this brave new world of rising inflation, a risk factor to consider are pricing changes. Very recently we have seen companies raising prices to balance increasing costs. Some organizations have even hired pricing gurus and developed sophisticated models for margin optimization. As part of these initiatives, it is very important to understand price elasticity, a term that economists use to describe how much more or less you will sell if the price changes.

One of the biggest mistakes when deploying these initiatives is to manage them only as standalone commercial projects that don’t all the stakeholders. Usually, the organization is very worried about margin and revenue, but doesn’t realize the impact of pricing on demand and inventory which affects productivity, working capital and customer service. Such impacts are hard to discern when looking at dollars only, since an increase in price could mask decreases in volume. It is important to understand unit changes as well.

Another impactful factor is called the bullwhip effect. It refers to how small fluctuations upstream in the supply chain become magnified as they trickle downstream. Think about how a small change in consumer behavior could create the need for higher inventory at the distribution center which in turn causes the need of a big production ramp up, which then triggers an even bigger raw material requirement. All this to have enough product in the network to support a smaller end user need. The main reason behind this situation is lack of visibility across all the supply chain, long lead times and safety stocks required at different stages in the end-to-end process.

I remember experiencing a good example of this concept in action. I was in a demand review meeting where the forecast generated by the Sales team did not make sense. They seemed very optimistic. So, when the Supply Chain team questioned the data, the Sales team made the point that the numbers provided were aligned to what the customer was selling in aggregate at their stores. But this was not consistent with the actual sales orders that the customer was placing with us. After both areas went to talk to the customer, the mystery was solved. This was an inventory reduction initiative from the client side to improve their working capital towards their fiscal year end. It turned out that both functions were right, but some information was missing. Armed with a good understanding of the situation, it was only a matter of calculating when the target inventory level would be reached by our client and then using the customer sales data as the forecast.

Additional elements that can influence your demand are promotions, movements from the competitors or new products, where an educated guess is required. Gauging this impact can be very subjective and requires a good understanding of the assumptions that are being used to determine demand. When working with retailers I have seen very simple assumptions in place that work well, for example determining how many products per store are need for a new product launch or how many more for a promotion and then just multiplying per the number of stores. Simple solutions that are easy to understand can be powerful as they rally the team behind a determined course of action.

The Art Behind Demand Planning

While Supply Chain usually does the legwork for the demand planning process, Sales is often held accountable for the forecast. Then you have input from Marketing, Product, and Operations teams that needs to be taken into consideration. There is a fine line that must be navigated where everyone needs to understand clearly what their piece of the pie is. Understanding who owns what and aligning the plan across functions is where the art of demand planning lies.

Managing this cross functional discussion is critical. A framework that I have found useful is RACI. This is used in each part of the process to describe who is responsible for each part of the work. ‘R’ is for roles, i.e., who does what part of the process. ‘A’ is for accountable, i.e., who needs to be consulted. ‘C’ is for consulted, i.e., those stakeholders whose feedback you need. ‘I’ is for informed, i.e. who needs to be informed of decisions.

This way there are no gray areas or confusion as to who is the decision maker. A suitable illustration of this can be the demand plan itself. As mentioned above, Supply Chain is usually in charge of the science portion of demand planning and getting a solid foundation in the form of a good statistical forecast. This same function facilitates the cross-functional discussion with other areas to come up with the demand plan. This could lead you to believe that Supply chain is accountable for the forecast, when they are responsible. The actual function that is accountable for the forecast is most commonly Sales. This is because they own the amount in dollars and the mix of the right product that the organization intends to commercialize.

Ironically, one of the barriers that I have encountered in the past is having leaders that want their sales team in the field focused on selling, regardless of how much or which product they sell, and not behind a computer. However, it is very critical to review at least once a month which products (and how much) are moving in the market to understand if we are going down the right path.

Gaining Consensus

There are different types of plans that exist in the business and that can be used in demand planning. The more common ones are:

• Budget or Annual Operating Plan: Done once a year with a fixed set of assumptions.
• Financial Forecast or Forecast Quarterly Update: This is an attempt to update the budget numbers to changes in the environment and is updated every three months.
• Sales and Operations Unconstrained / Constrained Demand Plan: This is updated on a monthly basis and captures the most recent circumstances, new product launches, recent historic trends and any new market intelligence. The difference between the unconstrained and constrained plan is that the former includes any operational, production, or supply constraints that impact what can be sold.

You can compare a constrained forecast to last year, last month sales, to budget or even last month’s forecast. This helps the company to understand if there is a gap versus previous commitments or substantial changes over time. But it needs to be understood clearly that the monthly constrained forecast currently discussed is taking into consideration the most recent circumstances and assumptions.

A big opportunity in this process that I have encountered over the years is when different plans exist in the company. This shifts the focus out of the most recent constrained forecast and could even create a situation where it is not being discussed cross functionally. The symptoms can be easily seen in an organization when finance has their own numbers, manufacturing is using their own forecast, sales have their own plan, while supply chain is trying to make sense of all of those and using yet another set of numbers to drive procurement. This creates real havoc and confusion and causes people to be caught like a deer in the head lights when trying to explain certain situations with their own set of data.

Sometimes having multiple plans is required by the business. If each plan is understood and used the right way, it can serve a purpose. But sometimes, it can become the easy way to avoid the tough conversations that are needed cross functionally. From my perspective, the best practice is to align into one main plan allowing executives to build contingencies based on risk levels of certain initiatives, promotions or economic variables. But Sales, Operations and Supply Chain need to have only one plan to which all metrics are applied and that aligns with the lead time needed to procure and ship each product. A good way to exemplify this is when measuring the forecast accuracy of an item with ninety days of lead time. The amount that you bought today will be based on what the forecast was three months ago.

Components of the Art of Demand Planning

• Always start the demand planning meeting with a solid foundation in the form of a good statistical forecast. Then focus on exceptional, new events or variables outside the models that could skew history, like promotions, price changes or inventory level at your customer.
• Understand who is responsible for the demand planning process and who is accountable for the actual plan. The former is the ultimate decision maker, but the areas that are responsible and consulted can provide input and pushback when required.
• Find consensus around one plan that includes the most recent circumstances and market dynamics. Having one plan that is wrong (within reason) but that rallies and aligns all the organization behind it, is better than having multiple plans across different areas.
• Allow a safe place to have the tough cross-functional conversations that are needed, discuss the data and assumptions, and learn from the iterative demand planning process.

Learn the fundamentals of demand and supply planning at IBF’s Supply Chain Planning Boot Camp in Nashville, from August 9-11, 2023. You’ll learn best practices covering Materials, Resource and Work Center Planning, Inventory Planning, the Supply Review, Supply Chain Metrics and more. Includes 1-day S&OP and Demand Management workshop. Register your place.

It is that time of year when leadership starts talking about strategy. The result will be very nice-looking slides that will be discussed in every town hall for the next month. When you are looking at the presentation, two questions will go through your mind.

First, how relevant are these initiatives to the company? You’ll have your own thoughts the direction the company wants to go in and wonder how serious the enterprise is about it. You’ve seen before how leaderships fails to follow through on their  own guidelines or change direction for some reason or another, making it hard for people on the ground to understand what the priorities are.

The second question that arises is: What’s my role in this in strategy? This is where you try to understand the impact it will have in your area and what will be required of you to support these strategic objectives Translating impactful PowerPoint slides to what we actually do day-to-day is easier said than done.

COVID showed business leaders the importance of Supply Chain Management so there is usually a section that links the overarching business plan to this area in a way that helps everyone involved in the process. As planning professionals, this is where we should focus our attention and seek to understand not what good enterprise strategy looks like but what good supply chain strategy looks like.

The Link Between Business Strategy & Supply Chain

“Plans are worthless, but planning is everything”, said Dwight D. Eisenhower. What we do as Demand Planners will invariably fail to reflect reality perfectly but are nevertheless valuable—indeed critical— to responding to demand and changing marketplace dynamics.

During the strategy ideation stage, the main question that the planning organization needs to answer is how to create value—value for the customer, employees, and even for our suppliers. If this is done successfully, it will set your company apart from the rest. I will not delve into the details of how this is done, rather I will provide an example of how to link this into the value chain.

Supply chain is a pilar that supports some of the core business objectives. How does your company’s strategy allow you to compete in it’s chosen market? And what supply chain model will support it most effectively? Supply chain can create value for our customers either through consistent and reliable delivery, truly short lead times, or great pricing. It is important that we choose the supply chain model that best aligns with the value creation strategy. Some examples of these are if we produce to a forecast, make to stock, only manufacturing when an order is received, or even only start designing the product once the order is confirmed. The following is a real-life example.

Real Life Example of a Planning Model

This happened during my first experience working in supply chain as a Master Scheduler. I worked for a factory that produced brass goods and had a wide assortment of products that kept growing over time. The main idea was to be able to manufacture these products in a reasonable amount of time and with an optimized amount of working capital. So, the model we used was Late Configuration. The company offered a wide variety of finished goods but with a lot of commonalities at the component and subassembly level.

The intent of the Late Configuration model was to wait and produce at the latest point of differentiation possible. To accomplish this, for example, you create buffers before a color change, or before you assemble the product and add a different option like another handle or trim. This allows you to absorb some of the demand variability of the finish goods in a subassembly that is common to several products, thus smoothing some of the variation by netting out puts and takes in the ordering pattern across different SKUs.

At the same time, it lets you reduce lead time as you are not starting productions from scratch and it has an inventory benefit as well, since the valuation of a semi-finished product is less than the finish goods and has a lower storage cost.

Finally, the supply signals were based on a pull system, using Kanban. This meant that if there was no demand, production would not be trigged and inventory would be kept at a component level. Components were acquired based on a forecast due to the long lead times, being sourced from Asia. This meant that if demand dropped after you filled the pipeline of semifinished products, all the excess inventory would be accumulated at the component level, which costs less and is cheaper to store. Obviously, the tradeoff is that you need to flex capacity. Adjusting staffing was the main way to change the output.

Real Strategy Vs Pie in the Sky

At the end, any strategy that you choose will be different depending on how you are creating value for the stakeholders in your business. But there is a sure way to identify a real strategy versus a wish list. Look at the tradeoffs. If you see a statement where the organization wants to provide an elevated level of service with little to no inventory, long lead times from suppliers and at a low cost, then this might be a clue. The classic tradeoff example that comes to mind when discussing this topic is about three attributes in a product or process. You can be fast, good, or cheap—but you can only pick two. This helps clarify supply chain decisions in a quite a straightforward way.

If you consider the Late Configuration example from the previous section, the model helped you reduce inventory, align production to demand, and have a reasonable lead time. But if demand changed a lot, you would have idle resources at the shop floor, creating additional costs or manufacturing variances to the financial plan. Another strategy for the company in question would be to produce all the finished goods assortment per the forecast. This could optimize manufacturing costs, reduce set ups, and slightly reduce lead time but will increase inventory and storage costs due to the complexity in product mix.

Going From Strategy to Execution 

“Culture eats strategy for breakfast”, said Peter Drucker. This highlights that while strategy is critical, it requires buy in and support from the whole organization to bring it to life. Since Management by Objectives was introduced in the 1950’s, the intention of closing the gap between what needs to be done and what is executed has been a very intensive journey. The combination of academic research and practical approaches has yielded a few frameworks that we can use. The main idea behind these concepts is that metrics drive behaviors and these in turn create a culture of execution in the company.

So, the next logical step is to go from top-level guiding principles to long term objectives, zoom into what the annual operating plan will look like and, finally, link this to Key Performance Indicators (KPIs). This is a straightforward process, and there are several methodologies available, like the Hoshin Kanri matrix if you are a fan of the Toyota Production System, or a balanced score card if you prefer classical methods.

The important aspect is to understand which part of the high-level objectives your area will have a real impact on. Then the priority is to cascade the measurements that are important for the organization in general into specific metrics that your department will own and deliver. In my experience, this is a terrific opportunity to spend some time together with your team (offsite to avoid distractions) and talk about how the supply chain organization creates a positive impact in the company and how we can measure it. At the same time, you can combine this with some team building activities to create relationships conducive to the development of a high-performance team.

A widely used method to define and deploy objectives is SMART Goals (Specific, Measurable, Attainable, Relevant and Time-bound). A recent trend, which is now one of my favorites, is FAST Goals (Frequently-discussed, Ambitious, Specific and Transparent), created by Don Sull from MIT. The main components of the former are intensive communication and stretch targets; both are key factors in developing the necessary culture. I will go back to my own experience to explain how this works in practice.

Several years ago, I was hired for a turnaround role as a Supply Chain Manager for a manufacturing site. The challenge was to increase the service level. The metric we had in place was on-time in-full (OTIF). After getting my head round their process, two things became clear. First, the bottleneck was at the finished goods warehouse. Second, Production was focusing on their own efficiency metrics. From an operations perspective, we had to add an additional shift and create Standard Operating Procedures (SOPs) to remove the constraint; this was very straightforward. However, from a scheduling perspective, we had to implement a daily cadence to discuss production deviations from the plan and understand the root causes. At the same time, we published the metric all over the plant, including the cafeteria.

At first, it was hard to stomach lunch while looking at an extremely low fill rate (OTIF). However, it generated a lot of internal discussion and a noticeably clear sense of priority. This created a tense but positive environment that supported the daily scheduling meeting and finally the process enabled the team to change the priority to mix over volume and hitting committed dates versus reducing the amount of set ups at the plant. After a few weeks of this, the metric started taking off and since it was very visible all over the plant, it generated a positive feedback loop that helped gather engagement from everyone and changed a very defeatist environment into one where everyone wanted to participate and contribute.

This allowed us to move the meeting cadence from daily to weekly and it became part of the operational review and culture of the plant. From this example you can see the importance of frequent communication and how a prominent level of transparency around metrics helps link the strategy directly into the culture of the organization.

One last word of caution when deploying FAST Goals: It is extremely critical that when reviewing stretch targets that there is a range in place and that the incentive plans for leaders are set up in tiers, so you can recognize ‘good’ and really reward the achievement of ambitious goals.

The Framework in a Nutshell

• Identify the role that Supply Chain plays in the bigger picture and make sure that the model fits the strategy.
• Call out the tradeoffs in the supply chain strategy to clearly define priorities.
• Cascade the business strategy all the way down to metrics that will define what success looks like. This will generate visibility of the impact that supply chain has in the organization.
• Make sure that there is frequent discussion around the metrics and the current performance of the team. Recognize ‘good’ but really reward excellence.
• Make sure that these processes are incorporated into the culture of your team; this will enable the creation of a high-performance organization.

Next time you are sitting in the strategy town hall, make sure that you apply some of these ideas. This will change your perception of these meetings from pretty slides with all power and no point (pun intended) into meaningful ways to make a difference in your organization.

This article first appeared in the winter 2022 issue of the Journal of Business Forecasting. To access the Journal, become an IBF member and get it delivered to your door every quarter, along with a host of memberships benefits including discounted conferences and training, exclusive workshops, and access to the entire IBF knowledge library.