At the heart of the supply chain are both the prediction of future customer demand and the satisfaction of actual customer demand. Reconciling the "prediction" to the "actual" is a growing problem. Increasingly fickle customer buying behavior, rising global competition, ever-shortening product lifecycles, product proliferation, world economic events -- the factors influencing demand volatility are far too numerous to list. Let's face it, it's an unpredictable world.
An Accurate Forecast Does Not Reduce Demand Volatility
Many companies that struggle with demand volatility try to address the issue by implementing a statistical forecasting tool. Here is the problem: implementing a forecasting tool isn't going to reduce the volatility. In fact, nothing will address demand volatility other than the root causes listed above, which are often outside of the control of a manufacturer.
Demand volatility is an expression of how much the demand changes over time. Forecast accuracy is an expression of how well one can predict the actual demand (whether it happens to be volatile or not). Why is it important to make this distinction?
Consider chemical engineering, which is a mathematically rigorous field in which complex equations are used to predict the behavior of complex systems. The equations are precise and outcomes can be predicted to the umpteenth decimal point. No one questions the validity or accuracy of the equations used, though there is continued research to "improve" the equations.
The equations used in chemical engineering are sufficiently accurate to design chemical plants. When it comes to actually running a chemical plant, all sorts of control systems are placed around the equipment to ensure the plant operates in a "stable" manner. It is recognized that all sorts of factors, outside the direct control of the plant operator — outside temperature, exact chemical composition of infeed materials, etc. — require that the plant be run outside of the exact assumptions made during the design of the plant. All sorts of readings are taken from the equipment, including the outflows, and continuously fed to the control systems to ensure a desired result.
In other words, there is an implicit understanding that the plant will not operate exactly as planned because of assumptions made during the design phase that are not realized in the production phase. This is analogous to demand planning. A statistical forecast can be calculated precisely but that won't mean the result will be any more accurate. Why?
Behavior Is Anything but Predictable
In most cases, we assume systems to be highly predictable and their behavior can be described very precisely by a set of equations. When last did you hear of a lead-time or production rate described as approximate? An inherent assumption is that these are so-called deterministic systems. According to Encarta, the definition of deterministic is
Clearly, the second definition applies most closely to this discussion, though the first introduces a very important point; the element of free will, which leads to unpredictable behavior or volatility. Nearly all social and business systems are based upon the notion of predictable behavior, but behavior is anything but predictable.
Consider yet another analogy. A person is checking IDs at the entrance to a bar at a rate of one minute per ID. If one person arrives at the door each minute, you would assume the system is balanced, and it would be reasonable to say there would be never be a line. Wrong. The line will grow indefinitely. Reflecting on the deterministic definition, why is that? Because, once the person checking the ID gets behind, a line forms, and there is no way to catch up. It will still take about a minute to check IDs and people are still arriving about one per minute. To understand this, one must realize that if the person checking IDs sits idle because there are no IDs to check, the unused time is lost forever and cannot be put to productive use. So the available time to check IDs is actually less than a minute given that the intervals between people arriving will fluctuate. So, while demand may have been accurately predicted in this case, the volatility factor still affected the ability to sufficiently meet demand.
Stop Building a "Better" Plan
So what's this got to do with supply chains? Let's be honest, any demand plan input is an average (at best) or an estimate (at worst). Yet we spend enormous amounts of time and energy fine tuning advance planning systems to provide better results, to the point that the results are assumed to be more accurate than the input data. Forecasts can be converted to financial forecasts and reported to the closest dollar, when in reality they probably represented an estimate of revenue to the closest $100,000.
The value of accurate information is not in question. What is being questioned is the value of spending lots of time and effort to make the inputs very accurate when they are only ever going to be approximations because of the inherent uncertainty in demand.
Most importantly, having a more accurate forecast isn't going to remove the volatility from the demand. Having a more accurate forecast isn't going to help the supply side to deal with the volatility on the demand side. The supply side must still be agile and flexible to adjust to the demand changes. And regardless of the time and effort put into statistical forecasting, in a dynamic market with lots of volatility, the forecast will always be inaccurate.
So the question is where should one spend time and effort? Do you try and make the demand plan as accurate as possible, including all the input data, and forever analyze why the actuals didn't match the plan? Or do you accept that there is a lot of uncertainty in the supply chain and devise ways to manage and respond quickly and effectively to change? It's no longer just about making a better demand plan, but about establishing a capacity to manage demand volatility.
Managing Demand — A Capabilities Assessment
What is needed is more coordination and less optimization; more human insight and less machine thinking; more collaboration and less control.
Clearly, it is important to get a fairly good understanding of the predicted demand, but recognize that the uncertainty inherent in the supply chain will drown out the "accuracy" of any optimized result. Start from the assumption that any forecast is an estimation of demand that will never be realized exactly as planned. The same goes for the supply plan. Go ahead and plan — everyone needs a plan — but also put into place processes and systems to detect and respond quickly and effectively to the inevitable changes in demand.
Humans are far better at dealing with that uncertainty than machines, so demand management in the 21st century should be about giving humans tools where they can use their judgment, in the face of volatility and uncertainty, to make fact-based decisions that address the surprise and compromise innate to today's global and multi-tier supply chains. Compromise is inherent in making decisions across multiple tiers of the supply chain. Humans are very good at reaching a compromise provided they understand the consequences of their actions. Machines do not understand compromise — so use humans to reach compromise and machines to evaluate the consequences.
At the heart of delivering this is the technical architecture of the enabling solution. Companies looking to transition from demand planning to demand management must look for technology solutions that can deliver the following capabilities:
Visibility
Alerting
Collaboration
Evaluation and Resolution
Who Does This Apply To?
Establishing a demand management capability applies to any global enterprise serving a customer-driven market. If, in your industry, the customer is in control, then you need to be demand-driven. In such an environment, you can't plan the customer — you must respond to them.
Traditional "push-based" supply chains, driven by statistical forecasts, put the focus on supply management. Given today's consumer-driven marketplace, companies are adopting demand-driven supply chains with "pull-based" models that are more customer-focused, which means there is a clear and urgent requirement for tools that facilitate dynamic demand management.
Think Broader than Demand Management
Demand management is a strategic priority for customer-focused companies, so it stands to reason that it requires a strategic solution to enable the process and empower the people responsible for it. But it is important to note, a solution for demand management does not need to be — and, in fact, should not be — a point solution. If you refer to the capabilities above, these provide a platform for many supply chain applications.
Demand and supply chain management is, by its very nature, a multidisciplinary business process. For companies that want to reap the potentially significant operational and financial benefits — and remain competitive in today's cutthroat business environment — reassessing the traditional solution architecture is paramount. Our new era of surprise and compromise requires a technology platform that will enable comprehensive and integrated demand-supply planning, monitoring and response capabilities.
The Bottom Line
Industry-leading organizations have learned that what you can't plan, you must respond to. In short, the best possible plans are made, and then everything unravels when there is a last-minute or unexpected change that needs an immediate response.
Too many companies are still managing demand response by brute force. For too long this critical component of driving the business has been underserved by inadequate tools for visibility, analysis and collaboration. Only a 21st century technical architecture can provide the core capabilities required to satisfy these requirements.
Achieving excellence in responding to changing customer demands has become a priority challenge facing enterprises today and can represent the largest opportunity for companies to increase customer service, enhance margins and attain more predictable revenue across the entire value chain.
An Accurate Forecast Does Not Reduce Demand Volatility
Many companies that struggle with demand volatility try to address the issue by implementing a statistical forecasting tool. Here is the problem: implementing a forecasting tool isn't going to reduce the volatility. In fact, nothing will address demand volatility other than the root causes listed above, which are often outside of the control of a manufacturer.
Demand volatility is an expression of how much the demand changes over time. Forecast accuracy is an expression of how well one can predict the actual demand (whether it happens to be volatile or not). Why is it important to make this distinction?
Consider chemical engineering, which is a mathematically rigorous field in which complex equations are used to predict the behavior of complex systems. The equations are precise and outcomes can be predicted to the umpteenth decimal point. No one questions the validity or accuracy of the equations used, though there is continued research to "improve" the equations.
The equations used in chemical engineering are sufficiently accurate to design chemical plants. When it comes to actually running a chemical plant, all sorts of control systems are placed around the equipment to ensure the plant operates in a "stable" manner. It is recognized that all sorts of factors, outside the direct control of the plant operator — outside temperature, exact chemical composition of infeed materials, etc. — require that the plant be run outside of the exact assumptions made during the design of the plant. All sorts of readings are taken from the equipment, including the outflows, and continuously fed to the control systems to ensure a desired result.
In other words, there is an implicit understanding that the plant will not operate exactly as planned because of assumptions made during the design phase that are not realized in the production phase. This is analogous to demand planning. A statistical forecast can be calculated precisely but that won't mean the result will be any more accurate. Why?
Behavior Is Anything but Predictable
In most cases, we assume systems to be highly predictable and their behavior can be described very precisely by a set of equations. When last did you hear of a lead-time or production rate described as approximate? An inherent assumption is that these are so-called deterministic systems. According to Encarta, the definition of deterministic is
de•ter•min•is•tic [ di tùrmi nístik ]
1. relating to determinism: relating to the doctrine or belief that everything, including every human act, is caused by something and that there is no real free will
2. of knowable outcome: having an outcome that can be predicted because all of its causes are either known or the same as those of a previous event
Clearly, the second definition applies most closely to this discussion, though the first introduces a very important point; the element of free will, which leads to unpredictable behavior or volatility. Nearly all social and business systems are based upon the notion of predictable behavior, but behavior is anything but predictable.
Consider yet another analogy. A person is checking IDs at the entrance to a bar at a rate of one minute per ID. If one person arrives at the door each minute, you would assume the system is balanced, and it would be reasonable to say there would be never be a line. Wrong. The line will grow indefinitely. Reflecting on the deterministic definition, why is that? Because, once the person checking the ID gets behind, a line forms, and there is no way to catch up. It will still take about a minute to check IDs and people are still arriving about one per minute. To understand this, one must realize that if the person checking IDs sits idle because there are no IDs to check, the unused time is lost forever and cannot be put to productive use. So the available time to check IDs is actually less than a minute given that the intervals between people arriving will fluctuate. So, while demand may have been accurately predicted in this case, the volatility factor still affected the ability to sufficiently meet demand.
Stop Building a "Better" Plan
So what's this got to do with supply chains? Let's be honest, any demand plan input is an average (at best) or an estimate (at worst). Yet we spend enormous amounts of time and energy fine tuning advance planning systems to provide better results, to the point that the results are assumed to be more accurate than the input data. Forecasts can be converted to financial forecasts and reported to the closest dollar, when in reality they probably represented an estimate of revenue to the closest $100,000.
The value of accurate information is not in question. What is being questioned is the value of spending lots of time and effort to make the inputs very accurate when they are only ever going to be approximations because of the inherent uncertainty in demand.
Most importantly, having a more accurate forecast isn't going to remove the volatility from the demand. Having a more accurate forecast isn't going to help the supply side to deal with the volatility on the demand side. The supply side must still be agile and flexible to adjust to the demand changes. And regardless of the time and effort put into statistical forecasting, in a dynamic market with lots of volatility, the forecast will always be inaccurate.
So the question is where should one spend time and effort? Do you try and make the demand plan as accurate as possible, including all the input data, and forever analyze why the actuals didn't match the plan? Or do you accept that there is a lot of uncertainty in the supply chain and devise ways to manage and respond quickly and effectively to change? It's no longer just about making a better demand plan, but about establishing a capacity to manage demand volatility.
Managing Demand — A Capabilities Assessment
What is needed is more coordination and less optimization; more human insight and less machine thinking; more collaboration and less control.
Clearly, it is important to get a fairly good understanding of the predicted demand, but recognize that the uncertainty inherent in the supply chain will drown out the "accuracy" of any optimized result. Start from the assumption that any forecast is an estimation of demand that will never be realized exactly as planned. The same goes for the supply plan. Go ahead and plan — everyone needs a plan — but also put into place processes and systems to detect and respond quickly and effectively to the inevitable changes in demand.
Humans are far better at dealing with that uncertainty than machines, so demand management in the 21st century should be about giving humans tools where they can use their judgment, in the face of volatility and uncertainty, to make fact-based decisions that address the surprise and compromise innate to today's global and multi-tier supply chains. Compromise is inherent in making decisions across multiple tiers of the supply chain. Humans are very good at reaching a compromise provided they understand the consequences of their actions. Machines do not understand compromise — so use humans to reach compromise and machines to evaluate the consequences.
At the heart of delivering this is the technical architecture of the enabling solution. Companies looking to transition from demand planning to demand management must look for technology solutions that can deliver the following capabilities:
Visibility
- Without a doubt, the greatest technical challenge is getting access to the right data on which decisions can be made. This wasn't as much of an issue in the days of integrated companies with a single enterprise resource planning (ERP) system, but in today's outsourced world, getting data from many different ERP systems, including those at contract manufacturers, suppliers and even customers, is a matter of necessity.
- True and valuable visibility is only achieved when there is fully automated integration of demand and supply data from multiple sources — including customer relationship management (CRM), ERP, demand planning tools, supply chain management (SCM) tools, spreadsheets and point-of-sale data — to provide a single integrated view of forecast, sales orders, inventory and supply.
Alerting
- A must-have capability is accurate and timely demand sensing to quickly understand demand shifts.
- Exception-based alerts to projected misalignments in demand and supply (e.g., forecast change beyond tolerance, late demand impacting quarter-end, supply disruptions, late customer orders, etc.) can enable demand managers and order fulfillment staff to quickly realign supply and/or engage manufacturing operations as needed.
Collaboration
- Resolution of an issue requires multiple people to collaborate in order to reach a consensus decision.
- Only human judgment applied through collaboration and supported by hard facts can allow teams to evaluate several innovative ways of resolving the issues at hand and reach a viable solution that best serves the interests of the company, customers and other stakeholders in the supply chain.
Evaluation and Resolution
- While it is important that all stakeholders can collaborate and evaluate alternative ways of resolving a supply chain issue, equally important is that they do this is in a manner that is consistent with both financial and operational corporate objectives. This requires an ability to evaluate the impact of all decisions (prior to execution) against the achievement of defined performance targets.
- Therefore, there is a need for decision support that drives profitable responses through shaping demand and allocating finished goods supply as appropriate.
Who Does This Apply To?
Establishing a demand management capability applies to any global enterprise serving a customer-driven market. If, in your industry, the customer is in control, then you need to be demand-driven. In such an environment, you can't plan the customer — you must respond to them.
Traditional "push-based" supply chains, driven by statistical forecasts, put the focus on supply management. Given today's consumer-driven marketplace, companies are adopting demand-driven supply chains with "pull-based" models that are more customer-focused, which means there is a clear and urgent requirement for tools that facilitate dynamic demand management.
Think Broader than Demand Management
Demand management is a strategic priority for customer-focused companies, so it stands to reason that it requires a strategic solution to enable the process and empower the people responsible for it. But it is important to note, a solution for demand management does not need to be — and, in fact, should not be — a point solution. If you refer to the capabilities above, these provide a platform for many supply chain applications.
Demand and supply chain management is, by its very nature, a multidisciplinary business process. For companies that want to reap the potentially significant operational and financial benefits — and remain competitive in today's cutthroat business environment — reassessing the traditional solution architecture is paramount. Our new era of surprise and compromise requires a technology platform that will enable comprehensive and integrated demand-supply planning, monitoring and response capabilities.
The Bottom Line
Industry-leading organizations have learned that what you can't plan, you must respond to. In short, the best possible plans are made, and then everything unravels when there is a last-minute or unexpected change that needs an immediate response.
Too many companies are still managing demand response by brute force. For too long this critical component of driving the business has been underserved by inadequate tools for visibility, analysis and collaboration. Only a 21st century technical architecture can provide the core capabilities required to satisfy these requirements.
Achieving excellence in responding to changing customer demands has become a priority challenge facing enterprises today and can represent the largest opportunity for companies to increase customer service, enhance margins and attain more predictable revenue across the entire value chain.
