Everand Logo

Welcome to Everand!

  • Discover millions of ebooks, audiobooks, and so much more with a free trial

    Only $11.99/month after trial. Cancel anytime.

    Improving Forecasts with Integrated Business Planning: From Short-Term to Long-Term Demand Planning Enabled by SAP IBP
    Improving Forecasts with Integrated Business Planning: From Short-Term to Long-Term Demand Planning Enabled by SAP IBP
    Improving Forecasts with Integrated Business Planning: From Short-Term to Long-Term Demand Planning Enabled by SAP IBP
    Ebook626 pages4 hours

    Improving Forecasts with Integrated Business Planning: From Short-Term to Long-Term Demand Planning Enabled by SAP IBP

    By Ganesh Sankaran, Federico Sasso, Robert Kepczynski and Alessandro Chiaraviglio

    ()

    Read preview

    About this ebook

    This book provides both a broad overview of the forecasting process, covering technological and human aspects alike, and deep insights into algorithms and platform functionalities in the IBP toolbox required to maximize forecast accuracy. Rich in technical and business explanations, it addresses short-, medium- and long-term forecasting processes using functionalities available in demand planning and demand sensing.

    There are also several theoretical concepts underpinning the algorithms discussed; these are explained with numerical examples to help demystify the IBP forecasting toolbox. Beyond standard procedures, the book also discusses custom approaches (e.g. new segmentation criteria, new outlier detection and correction methods) and new methods (e.g. the use of Markov chains for forecasting sporadic demands), etc. It subsequently benchmarks common practices using these innovative approaches and discusses the results. As measurement is an important precondition for improvement, an entire chapter is devoted to discussing process improvement and value using the Six Sigma methodology. In closing, the book provides several useful tips and tricks that should come in handy during project implementation. 

    LanguageEnglish
    PublisherSpringer
    Release dateMar 5, 2019
    ISBN9783030053819
    Improving Forecasts with Integrated Business Planning: From Short-Term to Long-Term Demand Planning Enabled by SAP IBP

    Related to Improving Forecasts with Integrated Business Planning

    Titles in the series (100)

    View More
    View More

    Related ebooks

    Sales & Selling For You

    View More
    View More

    Related podcast episodes

    Related articles

    • Article

      Q&A

      May 1, 2023

      Describe the capability that companies like Netflix, UPS, Amazon and Caesars Entertainment have in common. These are all leading firms in their industries with respect to leveraging analytics as a source of competitive advantage. We now have so much

      7 min read

    • Article

      How To Lead The Market In CX

      Feb 11, 2018

      Great customer experiences don’t happen by accident. They require a vision, a plan and flawless cross functional execution. In Avanade’s ‘Is your customer experience making an impact or not?’ this plan is broken down into a three-step ‘CX Trek’ of di

      3 min read

    • Article

      Change Sustainability Your ROI Health Check

      Mar 28, 2019

      Change sustainability programs are as unique as each of the workplace projects themselves. They can be developed either to dovetail into the end of a change program where businessas-usual kicks in, or they can be integrated into a prototype rotation

      5 min read

    • Article

      Human-centred Design

      Dec 23, 2018

      There was a recent report in The Sydney Morning Herald that was very misleading in its representation of the discipline of ergonomics. Citing the testimony of an authority on back pain, Sydney University professor, Chris Maher, the piece fundamentall

      4 min read

    • Article

      The Science Of Stock

      Feb 11, 2019

      The Science Of Stock
      8 min read

    • Article

      The Big Tech Boost

      Jan 5, 2024

      The Big Tech Boost
      5 min read

    • Article

      Customer-centric From Its Core

      Sep 16, 2018

      What’s been the biggest change you have seen in your career? I think one of the biggest things I am seeing recently is data becoming more and more important in how businesses operate and how they deliver customer experiences. Even a few years ago whe

      2 min read

    • Article

      Interviewing With Bots

      Oct 8, 2021

      imagine that your next job interview is with an artificial intelligence (AI) recruiting platform. It is a virtual meeting and the computer-generated person on your screen looks as life-like as you could imagine. It displays all the emotions and facia

      3 min read

    • Article

      Analytics Adolescent Or Mature Metrics Master?

      Jun 25, 2020

      An analytics company that specialise in advanced web analytics, Absolute Analytics, operates on the premise that every business has the right to actionable data and measurable results. From audit stage and set-up to training and reporting, these Alph

      4 min read

    • Article

      Fadi Naoum

      Mar 21, 2024

      Fadi Naoum
      3 min read

    • Article

      Getting The edge

      Feb 25, 2021

      Getting The edge
      7 min read

    • Article

      The Future Of Cannabis Data

      Jan 10, 2024

      The Future Of Cannabis Data
      3 min read

    • Article

      Smart Strategies

      Sep 22, 2021

      Smart Strategies
      3 min read

    • Article

      AGILE: It Is All About Trust

      Oct 22, 2019

      AGILE: It Is All About Trust
      6 min read

    • Article

      Data In A Digital World

      Sep 23, 2019

      Data In A Digital World
      3 min read

    Related categories

    Reviews for Improving Forecasts with Integrated Business Planning

    0 ratings

    0 ratings0 reviews

    What did you think?

    Tap to rate

    Review must be at least 10 words

      Book preview

      Improving Forecasts with Integrated Business Planning - Ganesh Sankaran

      © Springer Nature Switzerland AG 2019

      Ganesh Sankaran, Federico Sasso, Robert Kepczynski and Alessandro ChiaraviglioImproving Forecasts with Integrated Business PlanningManagement for Professionalshttps://doi.org/10.1007/978-3-030-05381-9_1

      1. Introduction

      Ganesh Sankaran¹  , Federico Sasso²  , Robert Kepczynski³   and Alessandro Chiaraviglio⁴  

      (1)

      Ingolstadt, Germany

      (2)

      Deloitte Consulting AG, Zurich, Switzerland

      (3)

      Rothrist, Switzerland

      (4)

      Polytechnic University of Turin, Turin, Italy

      Ganesh Sankaran (Corresponding author)

      Email: gmsankaran@gmail.com

      Federico Sasso

      Email: fsasso@deloitte.ch

      Robert Kepczynski

      Email: robkep0271@gmx.ch

      Alessandro Chiaraviglio

      Email: alessandro.chiaraviglio@gmail.com

      1.1 The What and Why of Forecasting

      This book is part of our Integrated Business Planning series with a special focus on how to make short, medium and long-term forecasting processes more effective and efficient through the use of rigorous analytical tools and techniques. We explain how we can walk away from the one-size-fits-all thinking to more tailored approaches for forecasting enabled by SAP IBP for demand. Let’s start by taking a step back and defining what forecasting is and why it is such a crucial process for any company.

      The organizational structures and capabilities that the companies can build to successfully navigate in uncertainty play a pivotal role in its success on the market. Such capabilities are particularly relevant when it comes to the demand planning and forecasting process. As uncertainty continues to grow in its fame with rapid change becoming the new normal, practitioners need effective tools to navigate the choppy waters of uncertainty. Forecasting is one such tool. Forecasting done well can serve as a steadying force—it can be seen as a stand-in or a proxy for reality. The better the forecast, the more ready is a practitioner when planning meets reality.

      Forecasting is the art and science of making predictions about the future. Science, because of the quantitative rigor involved in making extrapolations on the basis of relevant historical data. It is said that history doesn’t just repeat, it rhymes. The ability to see patterns and tell signal and noise apart calls for more than just science. This is one among several aspects of forecasting that justifies the expression the art of forecasting. Sometimes, it is better to not forecast and instead bide one’s time until more information becomes available. Daniel Kahneman in his influential book Thinking, fast and slow quotes comedian Danny Kays as having said of someone …her favorite sport is jumping to conclusions (Kahneman, 2011).

      To expand on this, one could say if it were an Olympic sport, it would probably be the one most fiercely fought. To avoid the trap of jumping to conclusions with not enough data is also something of an art form. It is also essential to, as Paulo Saffo puts it, be open to the full range of possibilities instead of a limited range of illusory certainties (Saffo, 2007). In this sense, the forecaster needs to not lose sight of the impact of her or his acts of commission and omission on decisions that will be made on the basis of the forecast. A keen understanding of the tools in the forecasting toolbox is therefore important in order of the key people involved in the process to not just be trusting bystanders, but active participants (Saffo, 2007).

      Let’s question the need for forecasting. If only customer order lead times can be negotiated to be at least as long as the total replenishment lead time (see Figs. 1.1 and 1.2.), most tactical and operational decisions can be made on the basis of real demands and the analog world of statistical forecasting would have limited use. Unfortunately (for companies and fortunately for customers), in this hypercompetitive, on-demand world, decisions have to be hedged on forecast. Therefore, the quality of forecast will directly or indirectly determine the quality of decisions and, consequently, the achievement of business goals. It is altogether fitting that demand forecast accuracy has pride of place in Gartner’s metrics hierarchy and sits atop the pyramid (more on this in the chapter on performance management)

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig1_HTML.png

      Fig. 1.1

      Order lead time versus total replenishment lead time

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig2_HTML.png

      Fig. 1.2

      Reality for most companies

      It is therefore absolutely essential to establish a sound basis for decision making. Sound basis for decision making just about summarizes the role of statistical forecasting. It is an extremely crucial cog in the wheel of demand management. It is perhaps the only unbiased input into the demand management process. Why? Machines don’t show biases, at least not today. With AI developing as fast as it is, who knows? Levity aside, statistical forecasting lends itself well as the first input, and an objective one, to the overall demand management process, which can then be enriched with other inputs based on more forward-looking factors and actions.

      In cognitive psychology, the concept of anchoring is well known and researched. It is the idea that, when faced with decision making, we humans tend to stick onto the first piece of information and start making adjustments anchored to this piece of information. Statistical forecasts are that anchor. If done poorly, it can adversely impact all of the activities that follow. In other words, first impressions matter.

      There are essentially two basic types of forecasting: qualitative (reliance on judgmental factors) and quantitative (consisting of time series and causal methods). Quantitative methods are (and should be) steeped in science, whereas there is a heavy dose of art involved in qualitative methods. In this chapter, our focus will be primarily on quantitative methods—we will use the term statistical forecasting to refer to these.

      When it comes to forecasting models, one could take one of two approaches: Try to understand causal factors and interdependencies between relevant control variables that impact what is being estimated or simply focus on the estimate alone. The former requires answering the question why, but the latter is simply about what. One can intuitively recognize that answering why requires far more sophistication (and has far higher data requirements) than answering what. Models that concern themselves with why are called explanatory models—causal algorithms belong in this category. Time series models on the other hand are only concerned with what (see Fig. 1.3). We will discuss a conceptual framework for selecting an appropriate algorithm later in the book.

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig3_HTML.png

      Fig. 1.3

      Why and what of forecasting

      Regardless of the type of algorithm, forecasting has its basis on the assumption of continuity. It relies on algorithms to interpret historical time series data and extrapolate it into the future. Sometimes, additional forward-looking inputs such as weather data or economic indicators are used, but in its simplest form the assumption is that history will repeat itself in some shape or form and that historical data are a good representation of how things will play out in the future.

      1.2 Planning Types

      Having looked at what forecasting is and why it is needed, let’s delve into the planning types and their relationship to the overall forecasting process. There are three planning types, namely operational, tactical and long-term planning (Fig. 1.4). In operational planning, the forecasting horizon is very short and it is granularity very high; this is the reason why demand sensing, which emphasizes the use of automation thereby limiting manual enrichment, plays a vital role in predicting short-term future. Tactical planning should be supported by statistical forecasting and cover horizons typically up to 3 years. Finally, in long-term planning, the activity of predicting the future happens on highly aggregated levels and often in monetary terms. As one can see, each of the planning types imposes certain requirements and challenges on the forecasting process. Let’s explore these further in the following paragraphs.

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig4_HTML.png

      Fig. 1.4

      Planning types and its characteristics

      Demand sensing as part of operational planning will need to deal with data with high granularity, often scattered, often from multiple sources (for example, in the case of point of sales data). A use case on how to integrate point of sales data with a special SAP solution into SAP IBP called SAP DSiM is provided in another book in our SAP IBP series—Implementing Integrated Business Planning (Kepczynski et al., 2018).

      Within tactical S&OP, a demand planning process strongly supported by statistical forecasting is most commonly leveraged to make predictions. They deal with products that may have stable but at the same time also high seasonal or intermittent demand patterns. It is not simple to forecast products that are highly seasonal, e.g., sold few months in a year or intermittent. We will shed light on how to address this challenge in the course of the book.

      Long-term planning covers in some industries up to 10-year horizon. Demand predictions are often based on qualitative inputs provided by marketing, business planning and finance. Long-term macroeconomic situations, researches and development pipelines, appetite and funds for mergers and acquisitions and plans to extend manufacturing footprints play a key role in this process. Predicting what will happen in the markets on a global scale relies on data that is typically coarse-grained leading to plans that are on a high level of aggregation.

      Let us summarize different types of planning and forecasting and provide the definition below for IBP which we will use for the purposes of this book:

      integrated business planning is a business management process which aims

      at connecting long term, tactical and operational planning,

      on local (markets, sites), regional (incl. production sites) and global level,

      to assess risk and opportunities, verify assumptions

      and generate with cross functional collaboration

      a feasible integrated business plan in volume and value.

      It is fundamental for the efficiency and effectiveness of the demand predictions to focus, as a primary step, on predicting volumetric sales. As to the question of what to forecast, the answer is quite easy—you should forecast what the market demands.

      As companies sell products and services (and not money!), it is essential that the sequence of forecasting reflects this fact: That is, volume precedes value.

      Very often, companies mix sales forecasts (market demand), operational plans and sales targets. Sales forecast is a demand projection, provided with a set of environmental assumptions. Operational plan is a set of operational actions required to reach the sales forecast regardless of the outcome. Sales targets are sales goals established to motivate the sales and marketing staff (Mentzer Jr & Moon, 2004).

      1.3 Scope of This Book

      The primary focus of this book will be on the demand-driven and volumetric side of the forecasting processes covering short-, medium- and long-term horizons.

      As to tools and techniques, algorithmic support (statistical forecasting and sensing algorithms) will be our particular focus throughout this book.

      Statistical forecasting techniques for long-term planning often rely on models that are based on selected variables to determine the impact on long-term volumetric predictions; e.g., the change of gross income per capita will increase the buying power of people in a specific region or country.

      In tactical S&OP, the demand review process is a crucial step. We have learned that statistical forecasting can be performed in different ways depending on the demand patterns. We want to share how to design statistical forecasting processes and how to leverage the out-of-the-box features and the instruments of SAP IBP to build your own customized features. In Fig. 1.5, we can observe that the demand review process is composed of the qualitative inputs from sales, marketing and demand planning, the quantitative inputs and of the right amount and quality of demand planning and data scientist capabilities. However, not all the components of the tactical S&OP process steps will be investigated in this book, but only those highlighted in Fig. 1.5. You may have noticed that between the process steps, there is cross/multiplication sign; the reason for such a choice relies on the fact that we strongly believe that the ultimate performance and efficiency of the demand review process depend on all the elements, acting as a whole and not as standalone ones.

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig5_HTML.png

      Fig. 1.5

      Forecasting in tactical S&OP, scope of the book

      Along with the traditional mid to long-term forecasting processes, which are part of the tactical S&OP and the long-term planning process, many companies leverage the demand sensing techniques. Operational planning uses demand sensing to detect signals from the market to make changes, within the short-term horizon, in the deployment/distribution plans, late customization plans and sometimes to even tackle production. Demand sensing may be perceived as an optimization technique for the consensus forecast that acts within the short-term horizon. Eventually, combining the outputs of short-term forecasting with mid- to long-term forecasting can give you several improvements in responsiveness and can help to make your supply chain more agile. Figure 1.6 illustrates the demand sensing topics that will be discussed in this book.

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig6_HTML.png

      Fig. 1.6

      Forecasting in operational planning, demand sensing—scope of the book

      Let us briefly explain the SAP IBP technology that supports the above processes. In SAP IBP, SAP Integrated Business planning is a real-time supply chain planning solution built to profitably meet the future demand by optimizing the supply chain. Built natively on SAP HANA and deployed on the cloud, SAP IBP provides the flexibility, agility and performance to meet complex planning requirements of the next-generation supply chain. With an integrated planning covering long-term, tactical and operational-level planning and a unified integrated model covering sales and operations planning, demand planning, inventory optimization, response and supply, and control tower, SAP IBP provides a single platform for all the planning needs. Together with robust planning algorithms, real-time simulations, what-if analysis, dashboards and analytics, alerts, embedded social collaboration and data integration with external sources, SAP IBP is the state-of-the-art planning solution (Fig. 1.7).

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig7_HTML.png

      Fig. 1.7

      SAP IBP applications overview

      SAP IBP for demand will provide a key logic for forecasting in the long-term, tactical and operational horizons. IBP for demand in large and complex organizations with comprehensive processes could be supported with improved visibility and exception management embedded in SAP IBP for Supply Chain Control Tower. In addition, another element which improves the end result is a functionality which facilitates collaboration; this is done via SAP JAM.

      In this book, we present how to leverage out-of-the-box functionalities for the processes we will be covering. Besides, we will describe some special configurations that were built making use of the standard mechanisms available in SAP IBP, e.g., the possibility to add stored key figures, calculated key figures, definition of custom rules to manage copy and default rules, definition of custom disaggregation and aggregation, of alerts and visualizations in Excel UI or Web UI. Last but not least, since the main user interface of SAP IBP is Excel, the reader could also learn how to combine the strengths of Excel with SAP HANA in the various planning templates.

      In SAP for demand, there are the following key groups of functionalities:

      Statistical forecasting

      Segmentation

      Demand sensing

      Forecast error measurement (see Fig. 1.8).

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig8_HTML.png

      Fig. 1.8

      SAP IBP demand application scope of the book

      We explained how we can improve focus and leverage segmentation in the book Integrated Business Planning use cases (Kepczynski et al., 2018). In this book, we will focus on statistical forecasting and demand sensing functionalities.

      1.4 Statistical Forecasting Process Framework

      Doing statistical forecasting right requires forethought and preparation and calls for a formal process framework. One such framework is presented in Fig. 1.9. We will use this framework as our reference model. The first step define purpose will be discussed in this chapter, and the remaining steps will be addressed in Chap. 3.

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig9_HTML.png

      Fig. 1.9

      Statistical forecasting process model

      1.5 Purpose

      Purpose definition involves getting clarity on various aspects such as: Who or what is most impacted by the forecast? What is the time frame of validity for decisions taken on the basis of forecasts? This is connected to the decision phase (long term/tactical/operational) the decisions belong to. One also needs to articulate the consequences of inaccuracy. This requires considering the nature of decisions that will be made based on the forecast as it will help determine the desired level of accuracy. For example, cost of item-level inaccuracy tends to be lower compared to inaccuracy on a higher level of aggregation (say, product group). Similarly, inaccuracies in the short term are also less dire when compared to mid-/long-term inaccuracies as these decisions are revisited more frequently.

      It also needed to be careful to not fall into the trap of forecasting for its own sake. It helps to have an appreciation for factors of uncertainty impacting the subject that is being forecasted while being mindful of the diminishing marginal utility of investing additional time and resources to unmask uncertainty in an effort to improve accuracy. When Donald Rumsfeld famously spoke about known unknowns and unknown unknowns at one of his press briefings during the height of the Iraq war, he must have been talking about epistemic and aleatory uncertainty (Rumsfeld, 2002). Epistemic uncertainty refers to things that are knowable, but aleatory refers to things that fall into the category of impossible to know for certain. When forecasting, it is important to acknowledge that sometimes it is impossible to predict certain things and a better approach in such instances would be to assess the risk exposure and hedge against such uncertainty, for example, by using time, capacity or inventory buffers.

      Span of control is also an important aspect to consider early in the forecasting process. This determines what gets classified as external, therefore the object of forecasting, and what gets classified as internal (controllable) and therefore is the scope of planning and decision making based on predictions of external factors (say, sales). In relation to this, it is worth mentioning that one should avoid forecasting requirements that can be derived, that is dependent requirements. It is the independent requirement (external demands) that needs to be the focus of forecasting, but it does not mean companies do not predict dependent requirements, e.g., for their highly important raw materials (steel for white goods manufacturers, active ingredient for chemical companies, varieties for agriculture industry). Levels in a bill of material are like the nodes of a supply chain and are similarly susceptible to the bullwhip effect if individual levels are forecasted in isolation.

      There are essentially two types of forecasts—market-oriented and production-oriented (Silver, Edward, Pyke, David, & Peterson, 1998). A clear understanding of purpose and scope determines the type of methods that are appropriate. Market-oriented methods are more qualitative in nature and when quantitative tend to be more sophisticated and expensive—higher forecasting costs. They drive decisions regarding resource requirements (long-term scope). Production-oriented methods on the other hand drive decisions regarding acquiring and scheduling of resources (tactical and operational scope).

      As long-term decisions involve higher degrees of freedom (see Fig. 1.10) or, in other words, offer a broader scope for decision making involving alternate scenarios and as such requiring a number of factors to be considered (external as well as internal), they call for more sophistication. Since decisions on this level have the power to change the givens (e.g., new markets and new locations), the assumption of continuity (past is a good indicator of the future) falls apart sometimes, which in turn calls for more qualitative or causal methods for forecasting. As one gets closer to execution, there is a pressure of the decision space and simpler time series methods start to become a lot more useful for decisions involving higher granularities and increased frequencies.

      ../images/471578_1_En_1_Chapter/471578_1_En_1_Fig10_HTML.png

      Fig. 1.10

      Decision making, forecasting and planning scope

      With a better appreciation for how purpose shapes the focus of forecasting, let’s turn our attention in the next section to some key challenges in forecasting related to both technical and human aspects.

      1.6 Key Challenges of Forecasting

      1.6.1 Seasonality

      Uncertainty, complexity and volatility in supply chains are here to stay. A well-established process for detecting and correcting seasonal outlying events represents one of the first steps toward supporting the companies in the improvement of forecast accuracy and customer service levels and reducing process variances and working capital tied to inventory.

      In addition, companies deal with the increased number of SKUs in their portfolios, strong seasonal sales cycles, exceptional events and intermittent or erratic sales patterns. Activities such as data cleansing, outlier detection and correction and forecasting are not currently properly supported by viable and agile processes and would require a specialized set of skills/capabilities and right system functionalities. Outlier detection, correction and forecasting are often addressed with too much human subjective guessing; in addition, the lack of data analysis for supporting decisions does not help for a more effective management. Those challenges and symptoms are even more exposed if we look on products that are highly seasonal or offer limited life, often not repeatable. It all results in high forecast errors, high supply chain and operating costs, loss of sales and not optimized inventory levels. Obviously, the overall result is frequently the following: During the season in which a company is supposed to account for most of its revenues, it is actually reporting for most of its lost compared to the realistic potential!

      The demand or the sales of a given product exhibit seasonality when the underlying time series undergo a predictable cyclic variation depending on a specific time within the year. More commonly, seasonality refers to regular periodic fluctuations that recur every year with about the same timing and intensity. For example, some consumer habits, like festivals, days off, and celebration days, could bring to seasonal patterns.

      Furthermore, there are some patterns that can be addressed as quasi-seasonal. They constantly reveal themselves every year, but with a different cadence or with discrepancies among countries or even within the years as the mother day which varies from country to country or Eastern which changes from year to year.

      Almost all businesses have some form of seasonality since factors such as weather, holidays, special events, specific time of the year, fiscal end-year and back-to-school day can all contribute to some short-term or long-term spikes in demand (Metersky, 2003).

      It is known that seasonality can be caused by internal and external factors, and it can create enormous inefficiencies. The seasonal cycle, which is the smallest time period for the repetitive cycle, has a huge impact on supply chains and operations; furthermore, the seasonal cycle influences and disturbs the variance of the ordering quantity, the forecasting process, the variance of the lead time and the normal ordering processes (Cho & Lee, 2012).

      Seasonality is uncertain and inevitable. It can happen at any time of the year, and it creates often managerial headaches and dramatic consequences if it is not addressed with the right attention (Amy & Partridge, 2017). It requires effective decision making and fast response and reactiveness.

      As a consequence, understanding, identifying and reacting to seasonality are subjects of interest for almost all the businesses. Improvement toward this topic can lead to a substantial benefit in the bottom line profits as well as to a competitive advantage toward other companies or to an added value to be exploited. Seasonality is also one of the most frequently used statistical patterns to improve the accuracy of demand forecasts.

      Among the influencers of seasonality, a brief mention should be addressed to the rogue seasonality, also called unintended cyclic variability. It is an endogenous disturbance generated by a company internal processes such as inventory, production and forecasting control systems ((Shukla, Naim, & Thornhill, 2012). Unintended seasonality can propagate all along the supply chain causing instability and higher cost propagation. It is often present in all supply chains with different degrees of intensity, and it commonly manifests itself in the ordering process. The unintended cyclic variations are present only for make-to-stock businesses, and it affects more the FMCG and technology businesses. The consequences of an unintended seasonality on a business are the same as the ones mentioned before, with the particularity that a root cause analysis is much more difficult to be assessed as all the departments contribute to the propagation of the disturbances and errors.

      When seasonality is unmanaged, the usual objective of getting the right amount, at the right place, at the right moment may become even trickier (Amy & Partridge, 2017). Seasonal stress is felt all along the supply chains and distributions networks: manufacturers, retailers, third parties, transportation carries and infrastructure outposts. It happens that everyone peaks at the same time and often, to make things worse, the 40/50% of the revenue of the year is earned in that very short lapse of time. Real-life examples suggest us problems such as lack of truckload capacity, congestions, nervousness and inaccurate demand planning (Amy & Partridge, 2017).

      Dealing with seasonality is money consuming, visibility gets poor, expenses are run at the last minute, sales are often missed, and organization stress reaches the same high peaks of the demand (Amy & Partridge, 2017).

      The common outputs of seasonality are lost revenues, inaccurate demand forecasting, missed production schedule, ineffective transportation and excessive inventory management (Cho & Lee, 2013) resulting in sky roofing operating costs. The phenomenon can also potentially cause a mismatch between supply and demand, and it frequently increases the costs of inventory and stock-out (Cho & Lee, 2013).

      Seasonality is not only related to the customer side, but also to the supply side; fluctuations may concern raw materials and labor availabilities hitting also the long-term corporate strategies. As a consequence, whether seasonality comes from supply or demand, the variability creates always significant challenges for all the actors involved.

      The forecasting processes often suffer from seasonal products as they are difficult to predict; their demands are unstable and always changing in quantities. Estimating the seasonality implies several complications as the following (Seasonality Definition, 2017):

      Time series are short. The life span of most consumer goods does not exceed 3 or 4 years. As a result, for a given product, sales history offers on average very few points in the past to estimate each seasonal cycle.

      Time series are noisy. Random market fluctuations impact the sales and make the seasonality more difficult to isolate.

      Multiple seasonality is involved. When looking at sales at the store level, the seasonality of the product itself is typically entangled with the seasonality of the store.

      Other patterns such as trend or product life cycle also impact time series, introducing various sorts of bias in the estimation.

      In addition, seasonal forecasting is complicated also because it requires a cross-functional collaboration and a rich information sharing. Often, information sharing is identified as the solution to enhance the company understanding of business seasonality. Moreover, the use of an established collaboration and shared inputs can lead the analysis to a further step of comprehending the root causes of each season. It is considered that a 360 degree approach will deliver the right benefits; for example, analyzing the historical peaks of the past year may require an understanding of the marketing activity, the contingency of that period, the promotions launched over the past years, the competitors’ activities during that period and many other influencers. Collecting information requires the right technology and the right skill set of people, because not all the data are useful and only few of them are the drivers toward the solution. Advanced mathematical analysis like the failure mode, effects and criticality analysis (FMECA) and the fault tree analysis (FTA) can enhance capabilities to understand seasonality and its impact.

      Sometimes, seasonality can appear easy to understand and predict: For example, if a company sells an anti-pollen pill, it is expected to see a peak during the spring; however, understanding and giving a weight to the influencers of the

      Enjoying the preview?
      Page 1 of 1