What makes a probabilistic forecast?
What’s all the hoopla around the term “probabilistic forecasting?” Is it just a more recent marketing term some software vendors and consultants have coined to feign innovation? Is there any real tangible difference compared to predecessor “best fit” techniques? Aren’t all forecasts probabilistic anyway?
To answer this question, it is helpful to think about what the forecast really is telling you in terms of probabilities. A “good” forecast should be unbiased and therefore yield a 50/50 probability being higher or lower than the actual. A “bad” forecast will build in subjective buffers (or artificially depress the forecast) and result in demand that is either biased high or low. Consider a salesperson that intentionally reduces their forecast by not reporting sales they expect to close to be “conservative.” Their forecasts will have negative forecast bias as actuals will nearly always be higher than what they predicted. On the other hand, consider a customer that provides an inflated forecast to their manufacturer. Worried about stockouts, they overestimate demand to ensure their supply. Their forecast will have a positive bias as actuals will nearly always be lower than what they predicted.
These types of one-number forecasts described above are problematic. We refer to these predictions as “point forecasts” since they represent one point (or a series of points over time) on a plot of what might happen in the future. They don’t provide a complete picture because to make effective business decisions such as determining how much inventory to stock or the number of employees to be available to support demand requires detailed information on how much lower or higher the actual will be! In other words, you need the probabilities for each possible outcome that might occur. So, by itself, the point forecast isn’t probabilistic one.
To get a probabilistic forecast, you need to know the distribution of possible demands around that forecast. Once you compute this, the forecast becomes “probabilistic.” How forecasting systems and practitioners such as demand planners, inventory analysts, material managers, and CFOs determine these probabilities is the heart of the question: “what makes a forecast probabilistic?”
Normal Distributions
Most forecasts and the systems/software that produce them start with a prediction of demand. Then they figure out the range of possible demands around that forecast by making incorrect theoretical assumptions about the distribution. If you’ve ever used a “confidence interval” in your forecasting software, this is based on a probability distribution around the forecast. The way this range of demand is determined is to assume a particular type of distribution. Most often this means assuming a bell shaped, otherwise known as a normal distribution. When demand is intermittent, some inventory optimization and demand forecasting systems may assume the demand is Poisson shaped.
After creating the forecast, the assumed distribution is slapped around the demand forecast and you then have your estimate of probabilities for every possible demand – i.e., a “probabilistic forecast.” These estimates of demand and associated probabilities can then be used to determine extreme values or anything in between if desired. The extreme values at the upper percentiles of the distribution (i.e., 92%, 95%, 99%, etc.) are most often used as inputs to inventory control models. For example, reorder points for critical spare parts in an electrical utility might be planned based on a 99.5% service level or even higher. While a non-critical service part might be planned at an 85% or 90% service level.
The problem with making assumptions about the distribution is that you’ll get these probabilities wrong. For example, if the demand isn’t normally distributed but you are forcing a bell shaped/normal curve on the forecast then how can then the probabilities will be incorrect. Specifically, you might want to know the level of inventory needed to achieve a 99% probability of not running out of stock and the normal distribution will tell you to stock 200 units. But when compared to the actual demand, you come to find out that 200 units only filled demand entirely in 40/50 observations. So, instead of getting a 99% service level you only achieved an 80% service level! This is a gigantic miss resulting from trying to fit a square peg into a round hole. The miss would have led you to take an incorrect inventory reduction.
Empirically Estimated Distributions are Smart
To produce a smart (read accurate) probabilistic forecast you need to first estimate the distribution of demand empirically without any naïve assumptions about the shape of the distribution. Smart Software does this by running tens of thousands of simulated demand and lead time scenarios. Our solution leverages patented techniques that incorporate Monte Carlo simulation, Statistical Bootstrapping, and other methods. The scenarios are designed to simulate real life uncertainty and randomness of both demand and lead times. Actual historical observations are utilized as the primary inputs, but the solution will give you the option of simulating from non-observed values as well. For example, just because 100 units was the peak historical demand, that doesn’t mean you are guaranteed to peak out at 100 in the future. After the scenarios are done you will know the exact probability for each outcome. The “point” forecast then becomes the center of that distribution. Each future period over time is expressed in terms of the probability distribution associated with that period.
Leaders in Probabilistic Forecasting
Smart Software, Inc. was the first company to ever introduce statistical bootstrapping as part of a commercially available demand forecasting software system twenty years ago. We were awarded a US patent at the time for it and named a finalist in the APICS Corporate Awards of Excellence for Technological Innovation. Our NSF Sponsored research that led to this and other discoveries were instrumental in advancing forecasting and inventory optimization. We are committed to ongoing innovation, and you can find further information about our most recent patent here.
Smart Software VP of Research to Present at Business Analytics Conference, INFORMS 2022
Dr. Tom Willemain to lead INFORMS session “Dominating The Inventory Battlefield: Fighting Randomness With Randomness.”
Belmont, Mass., March 2022 – Smart Software, Inc., provider of industry-leading demand forecasting, planning, and inventory optimization solutions, today announced that Tom Willemain, Vice President for Research, will present at the INFORMS Business Analytics Conference, April 3-5, 2022, in Houston, TX.
Dr. Willemain will present a session on how next-generation analytics arms supply chain leaders in manufacturing, distribution, and MRO with tools to fight against randomness in demand and supply. During his session he will detail the following technologies:
(1) Regime change filtering to maintain data relevance against sudden shifts in the operating environment.
(2) Bootstrapping methods to generate large numbers of realistic demand and lead time scenarios to fuel models.
(3) Discrete event simulations to process the input scenarios and expose the links between management actions and key performance indicators.
(4) Stochastic optimization based on simulation experiments to tune each item for best results.
Without the analytics, inventory owners have two choices: sticking with rigid operating policies usually based on outdated and invalid rules of thumb or resorting to subjective, gut-feel guesswork that may not help and does not scale.
As the leading Business Analytics Conference, INFORMS provides the opportunity to interact with the world’s top forecasting researchers and practitioners. The attendance is large enough so that the best in the field are attracted, yet small enough that you can meet and discuss one-on-one. In addition, the conference features content from leading analytics professionals who share and showcase top analytics applications that save lives, save money, and solve problems.
About Dr. Thomas Willemain
Dr. Thomas Reed Willemain served as an Expert Statistical Consultant to the National Security Agency (NSA) at Ft. Meade, MD, and as a member of the Adjunct Research Staff at an affiliated think-tank, the Institute for Defense Analyses Center for Computing Sciences (IDA/CCS). He is Professor Emeritus of Industrial and Systems Engineering at Rensselaer Polytechnic Institute, having previously held faculty positions at Harvard’s Kennedy School of Government and Massachusetts Institute of Technology. He is also co-founder and Senior Vice President/Research at Smart Software, Inc. He is a member of the Association of Former Intelligence Officers, the Military Operations Research Society, the American Statistical Association, and several other professional organizations. Willemain received the BSE degree (summa cum laude, Phi Beta Kappa) from Princeton University and the MS and Ph.D. degrees from Massachusetts Institute of Technology. His other books include: Statistical Methods for Planners, Emergency Medical Systems Analysis (with R. C. Larson), and 80 articles in peer-reviewed journals on statistics, operations research, health care, and other topics. For more information, email: TomW@SmartCorp.com or visit www.TomWillemain.com.
About Smart Software, Inc.
Founded in 1981, Smart Software, Inc. is a leader in providing businesses with enterprise-wide demand forecasting, planning, and inventory optimization solutions. Smart Software’s demand forecasting and inventory optimization solutions have helped thousands of users worldwide, including customers at mid-market enterprises and Fortune 500 companies, such as Disney, Arizona Public Service, and Ameren. Smart Inventory Planning & Optimization gives demand planners the tools to handle sales seasonality, promotions, new and aging products, multi-dimensional hierarchies, and intermittently demanded service parts and capital goods items. It also provides inventory managers with accurate estimates of the optimal inventory and safety stock required to meet future orders and achieve desired service levels. Smart Software is headquartered in Belmont, Massachusetts, and can be found on the World Wide Web at www.smartcorp.com.
SmartForecasts and Smart IP&O have registered trademarks of Smart Software, Inc. All other trademarks are their respective owners’ property.
For more information, please contact Smart Software, Inc., Four Hill Road, Belmont, MA 02478.
Phone: 1-800-SMART-99 (800-762-7899); FAX: 1-617-489-2748; E-mail: info@smartcorp.com
Smart Software Celebrates 40 years
40 years of Innovation for Demand Forecasting, Inventory Planning, and Supply Chain Analytics
Belmont, MA, June 1, 2021 – Today marks the 40th anniversary for Smart Software, a leading innovator of demand planning, statistical forecasting, inventory management, and supply chain analytics software.
Company CEO, Greg Hartunian remarked “Our success is built on continuous innovation. Our mission follows the path that our founders initiated 40 years ago; we provide cutting-edge analytical solutions that help our customers maximize sales and minimize waste. We are enormously grateful to our customers who have given us their support, confidence, and trust. Thank you to our partner community of resellers and consultants who have mobilized our growth and shared their expertise with us. We are also indebted to our many employees, past and present, local and abroad, whose creativity and dedication have produced systems that are benefitting so many great companies worldwide.”
Smart, Hartunian, and Willemain was incorporated in June 1981 by Charles Smart, Nelson Hartunian, and Thomas Willemain, our visionary founders. The firm later incorporated as Smart Software, Inc in 1984 reflecting their shift from boutique consultancy to software. Over the years, their pioneering work produced the first-ever automatic statistical forecasting system for the personal computer, a patented APICS award-winning method for intermittent demand planning, and most recently a cloud-native probabilistic forecasting platform. All have produced major inventory cost reductions and service level improvements for our customers. To learn more about Smart Software’s roots and journey, please click here:
Smart Software Company History
“Smart gives us good information to work with. The service level planning method has led to productive conversations between sales and supply chain and given us a common ground from which we base our discussions. People are feeling comfortable with numbers, and through our S&OP process we’ve been able to create buy-in across the company.”
Rod Cardenas – Purchasing Manager, Forum Energy
“It was deployed as part of our implementation of a new centralized distribution model and highlighted significant blind spots in the original project plan. The accurate forecasts of stocking levels and SKU count provided fact-based data that allowed us to strategically phase the consolidation effort where warehouse space was at a premium.”
Eric Nelson – CPA, CMA. Manager, Parts Supply and Logistics. BC Transit
“Its easy for us to give suppliers information they never had before. Our suppliers can plan their production and work with their suppliers. That visibility has been invaluable. That’s where the real payoff will come. Not just reducing inventory or saving time on people managing the inventory but being more responsive to customers’ needs. To me, that’s the overarching benefit of this software.”
Bud Schultz – Vice President of Finance NKK Switches
SmartForecasts and Smart IP&O have registered trademarks of Smart Software, Inc. All other trademarks are their respective owners’ property.
For more information, please contact Smart Software, Inc., Four Hill Road, Belmont, MA 02478.
Phone: 1-800-SMART-99 (800-762-7899); FAX: 1-617-489-2748; E-mail: info@smartcorp.com
Smart Software VP of Research to Present at Business Analytics Conference, INFORMS 2021
Dr. Tom Willemain to lead INFORMS session on Generation of Probabilistic Time-Series Scenarios
Belmont, Mass., March 2021 – Smart Software, Inc., provider of industry-leading demand forecasting, planning, and inventory optimization solutions, today announced that Tom Willemain, Vice President for Research, will present at the 2021 Virtual INFORMS Business Analytics Conference from April 12 – 14.
Dr. Willemain will present a session on Probabilistic Time-Series Scenarios, and how such scenarios are used, evaluated, and automatically generated using the statistical bootstrap. Frequently, OR models supporting business decisions are feed on massive numbers of probabilistic scenarios depicting future operating conditions. For example, with business operating at ever-lower levels of aggregation and ever-higher frequencies, demand planning and inventory optimization now use models fueled by scenarios representing the randomness of product demand at a daily scale. Dr. Willemain will discuss how even trivial decision tasks such as operator education benefit from large numbers of realistic training scenarios.
As the leading Business Analytics Conference, INFORMS provides the opportunity to interact with the world’s leading forecasting researchers and practitioners. The attendance is large enough so that the best in the field are attracted, yet small enough that you can meet and discuss one-on-one. The conference features content from leading analytics professionals, who share and showcase top analytics applications that save lives, save money, and solve problems.
Furthermore, to cutting-edge analytics content, the virtual analytics conference recognizes and prioritizes the need for quality “face-to-face” interactions, networking, and collaboration in a virtual setting.
About Dr. Thomas Willemain
Dr. Thomas Reed Willemain served as an Expert Statistical Consultant to the National Security Agency (NSA) at Ft. Meade, MD, and as a member of the Adjunct Research Staff at an affiliated think-tank, the Institute for Defense Analyses Center for Computing Sciences (IDA/CCS). He is Professor Emeritus of Industrial and Systems Engineering at Rensselaer Polytechnic Institute, having previously held faculty positions at Harvard’s Kennedy School of Government and Massachusetts Institute of Technology. He is also co-founder and Senior Vice President/Research at Smart Software, Inc. He is a member of the Association of Former Intelligence Officers, the Military Operations Research Society, the American Statistical Association, and several other professional organizations. Willemain received the BSE degree (summa cum laude, Phi Beta Kappa) from Princeton University and the MS and Ph.D. degrees from Massachusetts Institute of Technology. His other books include: Statistical Methods for Planners, Emergency Medical Systems Analysis (with R. C. Larson), and 80 articles in peer-reviewed journals on statistics, operations research, health care, and other topics. For more information, email: TomW@SmartCorp.com or visit www.TomWillemain.com.
About Smart Software, Inc.
Founded in 1981, Smart Software, Inc. is a leader in providing businesses with enterprise-wide demand forecasting, planning, and inventory optimization solutions. Smart Software’s demand forecasting and inventory optimization solutions have helped thousands of users worldwide, including customers at mid-market enterprises and Fortune 500 companies, such as Disney, Siemens, Metro Transit, APS, and The American Red Cross. Smart Inventory Planning & Optimization gives demand planners the tools to handle sales seasonality, promotions, new and aging products, multi-dimensional hierarchies, and intermittently demanded service parts and capital goods items. It also provides inventory managers with accurate estimates of the optimal inventory and safety stock required to meet future orders and achieve desired service levels. Smart Software is headquartered in Belmont, Massachusetts, and can be found on the World Wide Web at www.smartcorp.com.
SmartForecasts and Smart IP&O have registered trademarks of Smart Software, Inc. All other trademarks are their respective owners’ property.
For more information, please contact Smart Software, Inc., Four Hill Road, Belmont, MA 02478.
Phone: 1-800-SMART-99 (800-762-7899); FAX: 1-617-489-2748; E-mail: info@smartcorp.com
