SCM Part 2 Forecasting AY 25-26 STUDENT VERSION.pdf

# Introduction to forecasting Forecasting is a crucial process for predicting future sales quantities, playing a vital role in supply chain management by influencing strategic and operational decisions [2](#page=2) [6](#page=6) [9](#page=9). ### 1.1 The essence of forecasting Forecasting fundamentally involves predicting what will be sold, when it will be sold, and in what quantities. It is an essential tool for managing demand and ensuring that supply chain operations are aligned with anticipated customer needs [19](#page=19) [6](#page=6). ### 1.2 The role of forecasting within supply chain management Forecasting is integrated into a three-level planning hierarchy: strategic, tactical, and operational [9](#page=9). * **Strategic Level:** Involves defining the supply chain network and service offering [9](#page=9). * **Tactical Level:** Includes demand planning, sales, and operations planning, focusing on product mix and service levels [9](#page=9). * **Operational Level:** Encompasses executing and releasing orders, managing purchase and sales orders, and programming operations such as production, distribution, and sales [9](#page=9). Forecasting supports critical operational decisions by informing: * **Reorder points:** Determining when to reorder inventory based on demand during the lead time and the need for safety stocks [20](#page=20). * **Order quantities:** Influencing how much to order [20](#page=20). * **Capacity needs:** Guiding decisions on warehouse and production capacity requirements [20](#page=20). * **Component needs:** Ensuring the availability of necessary components [20](#page=20). Companies with more accurate forecasts typically experience substantially less inventory and a much lower percentage of stockouts [21](#page=21). ### 1.3 Challenges in forecasting Forecasting the future is inherently challenging due to several factors: * **Market dynamics:** The marketing mix, the introduction of innovative products, and the attractiveness of profitable segments to competitors can shorten the product life cycle (PLC) and complicate forecasting [10](#page=10). * **Product life cycle stages:** * **Introduction:** Demand is highly uncertain, and supply may be unpredictable. Product availability is critical, and costs are often secondary [11](#page=11). * **Maturity and Decline:** Demand becomes more certain, and supply predictable. Margins decrease due to competition, and price becomes a significant factor [12](#page=12). * **Promotional activities:** Forecasting the success of promotions is difficult but crucial. Factors contributing to this difficulty include assessing promotion success, competitor promotions, customer behavior leading to "loading" (stockpiling), and sales representatives' preselling efforts [14](#page=14). ### 1.4 The impact of promotions on supply chains Promotions significantly impact supply chains, creating a "promotion effect" characterized by a baseline sales level, a promotion peak, and a subsequent promotion low [15](#page=15). * **Definition of terms:** * **Baseline:** Regular sales without promotions [15](#page=15). * **Promotion peak:** The sales volume during the promotion period [15](#page=15). * **Promotion low:** The sales volume immediately following the promotion [15](#page=15). * **Lift factor:** The ratio of promotion peak to baseline sales ($ \text{Liftfactor} = \frac{\text{promotion peak}}{\text{baseline}} $) [15](#page=15). * **Dip factor:** The ratio of promotion low to baseline sales ($ \text{Dipfactor} = \frac{\text{promotion low}}{\text{baseline}} $) [15](#page=15). * **Net effect:** The difference between the surface area under the promotion peak and the surface area under the promotion low ($ \text{Netto effect} = \text{surface P} - \text{surface D} $) [15](#page=15). * **Lead/Lag:** Refers to the timing of sales shifts around the promotion [15](#page=15). The promotion effect can be observed at both the consumer and retail levels, influencing the flow of goods from supplier to retailer (retail effect) and from retailer to consumer (consumer effect) [16](#page=16). * **Implications for the supply chain:** The supply chain must adjust its output, increasing it to meet the promotion peak and then decreasing it to align with the subsequent dip. A portion of the increased sales during a promotion is merely a shift in demand from a future period (forward buying) or a reduction in future sales (cannibalization) [17](#page=17). > **Tip:** Forecasting promotions is often addressed by working with predetermined volumes or contingents to mitigate supply chain losses from obsolete inventory [14](#page=14). * **Examples of promotional pressure impact on additional sales:** * Benchmark demand (no promotions): 0% additional sales [18](#page=18). * Low promotional pressure: 1-1.5% additional sales [18](#page=18). * Middle promotional pressure: 3-4% additional sales [18](#page=18). * High promotional pressure: 10-11% additional sales [18](#page=18). (Note: Percentages can be lower for specific retailers like Dominicks, and middle promotional pressure was already experienced in early rounds) [18](#page=18). ### 1.5 Fundamental aspects and SCM drivers The effectiveness of supply chain management is influenced by the product's stage in its life cycle [13](#page=13). * **Impact on investment decisions:** The life cycle affects decisions related to capacity, leasing versus buying, and overall investment strategies [13](#page=13). * **Operational impact:** The product life cycle influences operational modes, such as shift systems, and the flexibility to switch production on or off [13](#page=13). * **Outsourcing decisions:** The stage of the product life cycle can impact decisions about in-house production versus outsourcing [13](#page=13). * **Efficiency:** Concepts like Agility and LEAN are considered within the context of the product life cycle's impact on the supply chain [13](#page=13). > **Tip:** Understanding how SCM drivers change across the product life cycle is essential for effective forecasting and supply chain planning [11](#page=11) [12](#page=12) [13](#page=13). --- # Characteristics and components of forecasts This section outlines the essential qualities of a forecast, acknowledging its inherent inaccuracies and detailing the key elements that constitute its structure. ### 2.3 Characteristics of a forecast Forecasts are fundamental to all planning decisions within a supply chain serving as the basis for both push and pull processes. They influence a wide range of operational and strategic activities, including production scheduling, inventory management, aggregate planning, sales force allocation, promotional strategies, new product introductions, plant and equipment investments, budgetary planning, and workforce planning. While sales managers are typically responsible for forecasting due to their market insights, accountants may rely on extrapolation from past sales and general trends [25](#page=25) [26](#page=26). Several key characteristics define the nature and reliability of forecasts: * **Forecasts are always inaccurate and should include both the expected value and a measure of error:** Due to inherent uncertainties, forecasts should not be presented as absolute truths. Instead, they must incorporate a measure of potential deviation from the actual outcome. This allows for better risk assessment and contingency planning [30](#page=30). * **Long-term forecasts are usually less accurate than short-term forecasts:** The further into the future a forecast extends, the more variables and unforeseen events can influence the outcome, leading to reduced accuracy. Short-term forecasts are generally more reliable for immediate tactical decisions [30](#page=30) [32](#page=32). * **Aggregate forecasts are usually more accurate than disaggregate forecasts:** Forecasting demand at a higher level of aggregation (e.g., total product category or all colors of apples) is typically more accurate than forecasting at a detailed product level (e.g., red apples, green apples). This is because the errors tend to cancel out when aggregating [24](#page=24) [30](#page=30). * **Information distortion increases further up the supply chain:** Companies located further upstream in the supply chain often receive distorted information due to cumulative forecasting errors and demand variability from downstream partners [30](#page=30). * **Forecasts are more accurate when observing a portion of demand:** Actively learning about demand and basing forecasts on observed demand patterns can lead to more accurate predictions and informs strategies like "Accurate Response". The diagram illustrates the relationship between order placement, lead times, and selling periods, highlighting the temporal aspects of demand observation [30](#page=30). ### 2.4 Components of a forecast Forecasting involves identifying and quantifying the factors that influence future demand and understanding the relationship between these factors and anticipated demand. Key components and considerations within a forecast include [31](#page=31): * **Factors Influencing Future Demand:** Companies must identify these elements and their impact on future demand. These include: * Past demand patterns [31](#page=31). * The lead time for product replenishment [31](#page=31). * Planned advertising or marketing efforts [31](#page=31). * Planned price discounts or promotions [31](#page=31). * The prevailing state of the economy [31](#page=31). * Actions taken by competitors [31](#page=31). * **Time Horizons:** Forecasts are typically categorized by their time horizon, with different horizons serving distinct planning purposes: * **Short-term forecasts (less than 3 months):** These are used for tactical decisions such as production planning and often involve "frozen periods" where plans are fixed [32](#page=32). * **Medium-term forecasts (up to 1 year):** These are crucial for budgeting and business planning. Inaccurate medium-term forecasts can lead to either excess inventory (tying up working capital) if too optimistic, or missed market opportunities if too pessimistic [32](#page=32). * **Long-term forecasts (3 years and more):** The duration of long-term forecasts depends on the industry. They play a significant role in strategic board-level decisions, such as building new factories or planning workforce needs [32](#page=32). * **Levels of Aggregation:** Forecasts can be generated at various levels of detail and scope: * **Hierarchical Levels:** Forecasts can start at broad levels (e.g., international, industry) and progressively break down to more specific levels (e.g., national, company, individual product) [33](#page=33). * **Temporal and Geographical Breakdown:** Once aggregated forecasts are established, they are further broken down by season and geography, potentially down to the level of individual salespersons [33](#page=33). > **Tip:** Understanding the distinct purposes and implications of short-term, medium-term, and long-term forecasts is critical for effective supply chain planning. > **Tip:** Recognizing that aggregate forecasts are generally more accurate than disaggregate ones can inform strategic decisions on where to place trust in forecasting efforts. ### 2.5 Basic approach of a forecast A systematic approach to forecasting involves several key steps to ensure its effectiveness and integration within the supply chain [34](#page=34): 1. **Understand the objective of forecasting:** Clearly define why the forecast is being made and what decisions it will inform [34](#page=34). 2. **Integrate demand planning and forecasting throughout the supply chain:** Ensure that forecasting is not an isolated activity but is embedded within the broader demand planning process and communicated across all relevant functions [34](#page=34). 3. **Identify the major factors influencing the demand forecast:** Analyze key drivers related to demand (e.g., growth, seasonality), supply (e.g., source lead times), and product characteristics (e.g., variants, substitution possibilities) [34](#page=34). 4. **Forecast at the appropriate level of aggregation:** Determine whether forecasts should be made at the supplier or store level, or another relevant aggregation point [34](#page=34). 5. **Establish performance and error measures for the forecast:** Define metrics to evaluate the forecast's accuracy and timeliness [34](#page=34). --- # Forecasting techniques Forecasting techniques are crucial for predicting future demand and are broadly categorized into qualitative and quantitative methods [38](#page=38). ### 3.1 Qualitative techniques of forecasting Qualitative techniques are primarily subjective and rely on expert judgment and opinions. They are particularly useful for forecasting demand for new products or in new markets where historical data is scarce [39](#page=39). #### 3.1.1 Consumer/user survey method This method involves directly asking customers about their anticipated purchases during a forecast period. It is often conducted by the salesforce, especially in business-to-business (B2B) contexts, through face-to-face interactions. This approach is most valuable when dealing with a small number of users who can provide reasonably accurate purchase intentions [40](#page=40). #### 3.1.2 Panels of executive opinion (Jury Method) In this "top-down" approach, internal or external experts with industry knowledge are consulted. Each expert prepares and defends their forecast within a committee setting [40](#page=40). #### 3.1.3 Salesforce composite This "bottom-up" method involves individual salespeople creating product-by-product forecasts for their respective sales territories. These individual forecasts are then aggregated to form a company-wide forecast. A potential advantage is that salespeople may have fewer complaints if their compensation is linked to the forecast, although this could also lead to pessimistic forecasts [40](#page=40). #### 3.1.4 Product testing / Test marketing This technique involves introducing a product to a limited market to gauge consumer response and forecast its broader market success. Test markets can be standard, controlled, or simulated [38](#page=38) [41](#page=41). ### 3.2 Quantitative techniques of forecasting Quantitative techniques utilize historical data and mathematical models to generate forecasts. They are best suited for situations with stable demand patterns [39](#page=39). #### 3.2.1 Components of an observation Observed demand ($O$) can be decomposed into a systematic component ($S$) and a random component ($R$) [42](#page=42). $$ O = S + R $$ The systematic component represents the expected value of demand and can include: * **Level:** The current deseasonalized demand [42](#page=42). * **Trend:** The growth or decline in demand over time [42](#page=42). * **Seasonality:** Predictable fluctuations in demand that occur within a specific period (e.g., year, quarter, month). Reasons for seasonality can include periods in the year, public holidays, sales efforts, annual price increases, or new product launches [42](#page=42) [43](#page=43) [44](#page=44). * **Random component:** The portion of demand that deviates from the systematic part, representing unpredictable variations [42](#page=42). Forecast error is defined as the difference between the forecast and the actual demand. A good forecast error is one whose size is comparable to the random component of demand [42](#page=42). #### 3.2.2 Time series analysis Time series analysis involves analyzing a sequence of data points measured at successive, uniformly spaced time intervals. The primary variable in time series forecasting is time itself. These methods can be static (unchanging) or adaptive (updated after each demand observation). Time series forecasting breaks down historical data into its components: level, trend, seasonality, and error. A significant risk is placing too much emphasis on past events to predict the future, making it most useful in relatively stable markets [46](#page=46). ##### 3.2.2.1 Moving average The moving average technique is employed when demand exhibits no discernible trend or seasonality, meaning the systematic component is primarily the level. The level in period $t$ ($L_t$) is calculated as the average demand over the last $N$ periods [52](#page=52). $$ L_t = \frac{D_t + D_{t-1} + \dots + D_{t-N+1}}{N} $$ The forecast for the next period ($F_{t+1}$) is simply the current level ($L_t$), and this forecast is also projected for subsequent periods ($F_{t+n} = L_t$). After observing the actual demand for period $t+1$ ($D_{t+1}$), the level estimate is revised [52](#page=52): $$ L_{t+1} = \frac{D_{t+1} + D_t + \dots + D_{t-N+2}}{N} $$ This revised level then becomes the forecast for period $t+2$ ($F_{t+2} = L_{t+1}$). Moving averages smooth out annual sales figures and are not ideal for periods with sudden upturns or downturns [52](#page=52) [54](#page=54). * **Example Moving Average Calculation:** Given weekly demands for milk: $D_1 = 120$ L, $D_2 = 127$ L, $D_3 = 114$ L, and $D_4 = 122$ L. * Forecast for Period 5 using a four-period moving average: $$ F_5 = L_4 = \frac{D_4 + D_3 + D_2 + D_1}{4} = \frac{122 + 114 + 127 + 120}{4} = \frac{483}{4} = 120.75 \text{ L} $$ * Forecast error if actual demand in Period 5 is $125$ L: $$ E_5 = F_5 - D_5 = 120.75 - 125 = -4.25 \text{ L} $$ * Revised demand for Period 6, using $D_5 = 125$ L: $$ L_5 = \frac{D_5 + D_4 + D_3 + D_2}{4} = \frac{125 + 122 + 114 + 127}{4} = \frac{488}{4} = 122 \text{ L} $$ $$ F_6 = L_5 = 122 \text{ L} $$ ##### 3.2.2.2 Simple exponential smoothing This method is also used when demand lacks a discernible trend or seasonality, focusing solely on the level. The initial estimate of the level ($L_0$) is typically the average of all historical data. Exponential smoothing weights past observations with exponentially decreasing weights, unlike simple moving averages where weights are equal [56](#page=56). ##### 3.2.2.3 Holt’s model (Trend-Corrected Exponential Smoothing) Holt's model is appropriate when demand is expected to have both a level and a trend in its systematic component, but no seasonality. The systematic component is modeled as the sum of level and trend [57](#page=57). ##### 3.2.2.4 Winter’s model (Trend- and Seasonality-Corrected Exponential Smoothing) Winter's model is used when the systematic component of demand includes a level, a trend, and a seasonal factor. The systematic component is represented as the product of the level and trend, multiplied by the seasonal factor: $Systematic \, component = (level + trend) \times seasonal \, factor$ [58](#page=58). #### 3.2.3 Adaptive Forecasting Adaptive forecasting involves updating estimates of the level, trend, and seasonality after each demand observation, incorporating all new data into these estimates [49](#page=49). ##### 3.2.3.1 Steps in Adaptive Forecasting 1. **Initialize:** Compute initial estimates for the level ($L_0$), trend ($T_0$), and seasonal factors ($S_1, \dots, S_p$) [50](#page=50). 2. **Forecast:** Predict demand for period $t+1$ ($F_{t+1}$) [50](#page=50). 3. **Estimate error:** Calculate the forecast error ($E_{t+1}$) as the difference between the forecast and the actual demand ($F_{t+1} - D_{t+1}$) [50](#page=50). 4. **Modify estimates:** Update the estimates for the level ($L_{t+1}$), trend ($T_{t+1}$), and seasonal factor ($S_{t+p+1}$) based on the computed error ($E_{t+1}$) [50](#page=50). #### 3.2.4 Causal methods Causal forecasting methods establish a relationship between demand and other influencing factors such as economic indicators, interest rates, or price promotions [39](#page=39). #### 3.2.5 Simulation Simulation involves creating models that imitate consumer choices to understand the demand patterns that arise, often used for "what if" scenario planning [39](#page=39). ### 3.3 Forecasting and the Bullwhip effect While not detailed in the provided text, the context suggests that forecasting plays a role in understanding and potentially mitigating the bullwhip effect [35](#page=35). ### 3.4 Measures of forecast error and accuracy Several metrics are valuable for assessing forecast error and accuracy [59](#page=59). #### 3.4.1 Bias Bias represents the average deviation of the forecast from actual demand over a longer period. It indicates if the forecast consistently overestimates or underestimates demand [60](#page=60). #### 3.4.2 Mean Absolute Deviation (MAD) MAD measures the average magnitude of forecast errors, irrespective of their direction [59](#page=59). #### 3.4.3 Mean Squared Error (MSE) MSE calculates the variance of the forecast error, giving more weight to larger errors [59](#page=59). #### 3.4.4 Mean Absolute Percentage Error (MAPE) MAPE quantifies the average absolute error as a percentage of actual demand, providing a relative measure of error [59](#page=59). #### 3.4.5 Tracking Signal The tracking signal is used to monitor the cumulative forecast error and should ideally remain within certain bounds (e.g., less than 6) to indicate a stable forecasting process [59](#page=59). #### 3.4.6 Forecast Accuracy (FA) Forecast Accuracy (FA) is a calculated percentage, ranging from 0% to 100%, indicating how close the forecast was to actual demand. A perfect forecast has an FA of 100% [62](#page=62). The formula for Forecast Accuracy is: $$ FA_n = 1 - \frac{|E_n|}{D_n} \times 100\% $$ where $|E_n|$ is the absolute forecast error for period $n$ and $D_n$ is the actual demand for period $n$ [62](#page=62). * **Example Forecast Accuracy Calculation:** Given an estimated demand for period 5 of 120.75 L and an actual demand of 125 L, resulting in a forecast error of -4.25 L. $$ FA_5 = 1 - \frac{|-4.25|}{125} \times 100\% = 1 - \frac{4.25}{125} \times 100\% = 1 - 0.034 \times 100\% = 1 - 3.4\% = 96.6\% $$ *(Note: The provided document states 97% for the example which implies rounding or a slightly different calculation in their example context. The calculation here follows the formula precisely.)* [64](#page=64). #### 3.4.7 Forecast Accuracy Analysis Analyzing forecast accuracy involves using metrics to eliminate bias and generating reports that compare actual demand with forecasts across different levels and planning versions. Sorting by forecast error can help identify areas for improvement [61](#page=61). ### 3.5 Types of forecasting errors * **Over-forecasting:** Occurs when the forecast is higher than the actual demand [66](#page=66). * **Under-forecasting:** Occurs when the forecast is lower than the actual demand [66](#page=66). Both types of errors can lead to increased costs and are time-consuming to manage [66](#page=66). > **Tip:** Using multiple forecasting methods is often more effective than relying on a single technique [39](#page=39). > > **Tip:** When dealing with new products or markets, qualitative techniques are generally more appropriate due to the lack of historical data [39](#page=39). > > **Tip:** Quantitative time series methods are best suited for stable markets where historical patterns are likely to persist [46](#page=46). --- # Forecasting and the bullwhip effect This section examines how forecasting practices contribute to the bullwhip effect and discusses strategies for mitigation. ### 4.1 Understanding the bullwhip effect The bullwhip effect is a phenomenon in supply chains where demand variability increases as one moves upstream from the customer. Orders placed with suppliers tend to have a larger variance than the actual sales to the buyer, leading to information distortion and amplified variance propagation upstream [68](#page=68). ### 4.2 Causes of the bullwhip effect linked to forecasting Several forecasting-related practices can directly contribute to the bullwhip effect: #### 4.2.1 Demand signal processing * When a supply chain player updates its order-up-to-level based on its own demand forecast, the variance in its orders often exceeds the variance in the observed demand [70](#page=70). * This distortion is amplified when a player forecasts demand based on orders received from downstream, rather than observing the final customer demand [70](#page=70). * Larger lead times exacerbate the bullwhip effect caused by demand signal processing [70](#page=70). #### 4.2.2 Price variations and forward buying * Manufacturers offering promotions can incentivize retailers to engage in "forward buying," purchasing more than immediately needed in anticipation of future demand [71](#page=71). * This forward buying can artificially inflate demand signals. After the promotion ends, actual demand typically falls below the initial level for a period, creating further distortion in the demand information communicated upstream [71](#page=71). * This cycle leads to increased inventory costs for both retailers (holding inventory ahead of need) and manufacturers (preparing for demand surges) [71](#page=71). ### 4.3 Strategies for mitigating the bullwhip effect through forecasting Effective strategies focus on improving information flow and collaboration across the supply chain. #### 4.3.1 Information sharing Sharing information between supply chain partners can significantly reduce the bullwhip effect. In a two-level supply chain involving a manufacturer and a retailer, this includes sharing data such as [72](#page=72): * Demand [72](#page=72). * Inventory levels [72](#page=72). * Inventory policies [72](#page=72). * Promotion plans [72](#page=72). * Manufacturer's inventory and capacity [72](#page=72). #### 4.3.2 Demand sharing and cooperative forecasting Demand sharing, where a retailer shares its demand information with the manufacturer, offers several benefits: * **Benefits to the manufacturer:** * Reduced demand variance [73](#page=73). * Lower safety stock requirements [73](#page=73). * Reduced need for flexibility [73](#page=73). * Lower inventory smoothing costs [73](#page=73). * **Benefits to the retailer:** * Increased certainty of timely delivery and supply [73](#page=73). * Potential for reduced actual lead times if the retailer faces more stable demand due to information sharing (assuming the manufacturer does not have infinite capacity) [73](#page=73). To facilitate demand sharing, arrangements should be made where the retailer shares information, and the manufacturer shares cost savings through mechanisms like: * Vendor Managed Inventory (VMI) [74](#page=74). * Manufacturer-offered discounts to the retailer [74](#page=74). * Manufacturer efforts to reduce lead times [74](#page=74). #### 4.3.3 Vendor managed inventory (VMI) VMI can specifically help to lower the impact of the bullwhip effect [75](#page=75). #### 4.3.4 Promotion information sharing Retailers sharing their promotion plans with the manufacturer can further reduce the bullwhip effect. This information sharing is particularly beneficial in [76](#page=76): * Environments with high customer stockpiling (high promotion sensitivity), such as for products with long shelf lives [76](#page=76). * More competitive and less predictable environments with high demand variance [76](#page=76). Without shared promotion information, retail promotions may decrease profits for the manufacturer. Cooperative forecasting or forecast collaboration is a key concept underlying these mitigation strategies [72](#page=72) [73](#page=73) [74](#page=74) [76](#page=76). > **Tip:** The core idea behind mitigating the bullwhip effect through forecasting is to move away from independent, reactive forecasting at each stage of the supply chain towards a collaborative, integrated approach that shares actual demand data and future plans. --- # Demand planning and sales & operations planning (S&OP) Demand planning and Sales & Operations Planning (S&OP) are critical processes for aligning an organization's strategic goals with its operational execution by balancing supply and demand [78](#page=78) [90](#page=90). ### 5.1 Demand planning Demand planning is a process that combines various inputs, including statistical forecasts, customer information, sales and marketing insights, and promotions, to create a single operational forecast. A successful demand planning process requires a sound forecasting methodology coupled with the ability to respond quickly to changes. It emphasizes careful evaluation of all variables and inputs, with sales, marketing, and operations playing crucial roles through shared vision and ownership. Regular updates and flexibility are also key to adapting to evolving situations and the external environment [79](#page=79) [80](#page=80). #### 5.1.1 Fundamental capabilities and benefits of demand planning Effective demand planning incorporates customer, sales, marketing, and statistical forecasts into one operational forecast. This process allows for updates based on sales, marketing, or customer information, and tracks promotions and events. It also facilitates cross-functional coordination for new product introductions and maps key accounts to stocking locations. The benefits include improved customer service, reduced inventory and obsolescence (leading to fewer stockouts), increased inventory throughput, and decreased premium distribution costs [80](#page=80). #### 5.1.2 Elements of effective demand planning Key elements for effective demand planning include: * **Data modeling**: Understanding the components of demand, which can include levels, trends, seasonality, cycles, random variations, and events [82](#page=82). * **Promotion management**: Planning and tracking promotions, including their impact on historical data and future forecasts. This involves selecting promotion variants, defining key figures, products, planning versions, and start dates/periods [83](#page=83) [84](#page=84) [85](#page=85). * **Consensus forecasting**: Supporting the S&OP process by enabling the creation of multiple plans (e.g., product level for marketing, sales areas for sales, distribution centers for operations, business units for finance) and integrating them into a single consensus plan. This involves reconciliation and combination of plans at various levels. Consensus forecasting provides facts, such as history, events, and causal analysis, and ensures that adjustments can be translated across all company levels [86](#page=86) [87](#page=87). * **Product life cycle management**: Considering the introduction of substitute products and adjusting forecasts accordingly, such as accounting for launch ramp-ups for new products and baseline adjustments for older products [88](#page=88). * **Integration with sales & operations planning**: Linking demand planning directly with the broader S&OP framework. ### 5.2 Sales & Operations Planning (S&OP) Sales & Operations Planning (S&OP) is an integrated business management process designed to achieve focus, alignment, and synchronization among all organizational functions. It aims to balance supply and demand over a 6-12 month horizon. The core idea of S&OP is to forecast sales and production capabilities, take actions to balance supply and demand, examine the financial consequences of these actions against targets, and then initiate actions to bridge any gaps [91](#page=91) [97](#page=97). #### 5.2.1 Objectives of S&OP The primary objectives of S&OP are: 1. To balance supply and demand within the organization [96](#page=96). 2. To maintain correct inventory levels [96](#page=96). 3. To translate financial targets into operational plans [96](#page=96). 4. To align different departments [96](#page=96). 5. To facilitate proactive planning and anticipation of changing situations [96](#page=96). 6. To monitor performance [96](#page=96). #### 5.2.2 The S&OP process S&OP is often described as a five-step business planning process: 1. **Demand**: Moving from forecasting to demand shaping [98](#page=98). 2. **Supply**: Moving from capacity planning to supply network trade-offs and design [98](#page=98). 3. **Reconcile demand and supply**: Balancing the forecasted demand with the available or planned supply [97](#page=97) [98](#page=98). 4. **Reconcile with financial plans**: Aligning the operational plans with financial targets and budgets [98](#page=98). 5. **Sales and Ops Planning Meeting**: A forum for reviewing, decision-making, and generating a consensus plan, often involving "what if" scenarios rather than simple "yes/no" answers [98](#page=98). > **Tip:** S&OP meetings should be fact-based, not emotion-based, with clear ownership of decisions and formally linked operating and financial plans [99](#page=99). #### 5.2.3 Challenges and principles of S&OP Despite its seemingly simple concept, S&OP presents several challenges: * Discussions need to be fact-based rather than emotion-based [99](#page=99). * Clear ownership for each process element and decision is essential [99](#page=99). * The operating plan must be formally linked by assumptions to a financial plan [99](#page=99). * Formal balancing of demand and supply across a rolling horizon is required [99](#page=99). * Gaps against targets must be identified, and action plans formulated considering relevant lead times [99](#page=99). * Trade-offs must be clearly articulated and evaluated for their commercial value and risks [99](#page=99). #### 5.2.4 Balancing capability and demand shaping S&OP involves a balancing act between capability response and demand shaping [100](#page=100). * **Capability response** includes strategies like postponement, improving transparency (e.g., Vendor Managed Inventory - VMI), designing for supply, implementing logistics policies, creating adaptive networks, adopting flexible manufacturing, and tying agility strategies to demand shaping [100](#page=100). * **Demand shaping** encompasses marketing programs, new product introductions, promotions, trade deals, sales incentives, price management, and supply shaping/run-out strategies [100](#page=100). #### 5.2.5 Behavior and collaboration in S&OP S&OP is as much about behavior and collaboration as it is about process. Effective S&OP involves open communication, accountability for forecast accuracy, and proactive problem-solving to close gaps between forecasts and targets. It moves away from a functional silo approach and firefighting towards a key decision-making forum where issues are addressed early, leading to efficient responses and anticipation . #### 5.2.6 Benefits of a well-functioning S&OP A successful S&OP process yields significant benefits: * Improved customer service through aligned, anticipated actions . * A simplified decision-making framework . * Alignment of organizational actions and rewarding them accordingly . * Enhanced visibility of information . * Empowerment, development, and motivation of people . * Clarity on roles and responsibilities . * Ownership of key inputs . #### 5.2.7 Inputs from different functions in S&OP Various departments contribute essential information to the S&OP process . * **Marketing**: Provides market intelligence, product strategy, and demand shaping input . * **Sales**: Contributes customer interface information, sales targets, and demand forecasts . * **Operations**: Shares production capacity, workforce availability, material constraints, and supply network information . * **Finance**: Provides capital availability, business plan targets, and financial consequences of proposed plans . * **Product Development**: Informs about new product introductions and life cycle management . * **Supply Chain Management**: Brings insights on planned safety stock, order quantities, and supplier constraints . #### 5.2.8 S&OP definition Sales and operations planning (S&OP) is defined as an integrated business management process where the executive team continually achieves focus, alignment, and synchronization across all organizational functions. This process involves an updated forecast that generates a sales plan, production plan, inventory plan, customer lead time (backlog) plan, new product development plan, strategic initiative plan, and the resulting financial plan. The frequency and horizon of planning depend on industry specifics, with shorter product life cycles and higher demand volatility requiring tighter S&OP. When executed well, S&OP enables effective supply chain management by routinely reviewing customer demand and supply resources, and re-planning quantitatively over an agreed rolling horizon, focusing on future actions and anticipated results . --- ## Common mistakes to avoid - Review all topics thoroughly before exams - Pay attention to formulas and key definitions - Practice with examples provided in each section - Don't memorize without understanding the underlying concepts

| Term | Definition | |------|------------| | Forecasting | The process of predicting future sales or demand, which serves as the basis for various planning decisions within a supply chain. | | Demand Planning | A process that combines statistical forecasts with sales and marketing intelligence to create an operational plan for meeting customer demand. | | Sales & Operations Planning (S&OP) | An integrated business management process that aligns sales forecasts with operational capabilities to balance demand and supply over a planning horizon. | | Bullwhip Effect | A phenomenon in supply chains where demand variability increases as orders move upstream from the customer to the manufacturer, leading to inefficient inventory levels and operational disruptions. | | Qualitative Techniques | Forecasting methods that rely on subjective judgment, opinions, and intuition, often used for new products or situations with limited historical data. | | Quantitative Techniques | Forecasting methods that use historical data and mathematical models to predict future demand. | | Time Series Analysis | A quantitative forecasting method that analyzes historical data points collected over time to identify patterns like level, trend, and seasonality, and extrapolates these patterns into the future. | | Moving Average | A simple quantitative forecasting technique that calculates the average demand over a specified number of past periods to predict future demand, smoothing out fluctuations. | | Exponential Smoothing | A time series forecasting method that assigns exponentially decreasing weights to past observations, giving more weight to recent data when forecasting. | | Holt’s Model | An extension of exponential smoothing that accounts for both the level and trend in the systematic component of demand. | | Winter’s Model | A time series forecasting method that incorporates level, trend, and seasonality into its predictions, suitable for demand with predictable seasonal patterns. | | Forecast Error | The difference between the actual demand and the forecasted demand. | | Mean Absolute Deviation (MAD) | A measure of forecast error that calculates the average of the absolute differences between forecasted and actual values. | | Mean Squared Error (MSE) | A measure of forecast error that calculates the average of the squared differences between forecasted and actual values, giving more weight to larger errors. | | Promotion Effect | The impact of promotional activities on sales demand, often leading to temporary increases in demand followed by dips, and requiring careful forecasting. | | Liftfactor | A metric used to quantify the increase in sales during a promotion compared to the baseline sales without promotions. | | Dipfactor | A metric used to quantify the decrease in sales after a promotion period compared to the baseline sales. | | Forward Buying | A practice where customers purchase larger quantities of a product than they immediately need, often in anticipation of future price increases or to take advantage of current promotions. | | Cannibalization | A situation where a new product or promotion reduces the sales of an existing product from the same company. | | Aggregate Forecast | A forecast for a group of products or a larger time period, which is generally more accurate than disaggregate forecasts. | | Disaggregate Forecast | A forecast for individual products or specific time periods, which is typically less accurate than aggregate forecasts. | | Lead Time | The duration between the initiation of a process and its completion, such as the time between placing an order and receiving the goods. | | Safety Stock | Extra inventory held to mitigate the risk of stockouts due to demand or supply variability. | | Vendor Managed Inventory (VMI) | A supply chain strategy where the supplier is responsible for managing the customer's inventory levels, aiming to reduce inventory costs and improve availability. | | Consensus Forecasting | A forecasting process where input from various departments (sales, marketing, operations, finance) is gathered, discussed, and reconciled to produce a single, agreed-upon forecast. | | Product Life Cycle (PLC) | The stages a product goes through from its introduction to the market until its withdrawal, including introduction, growth, maturity, and decline, each with different demand characteristics. | | Demand Shaping | Strategies employed to influence customer demand, such as through marketing programs, promotions, pricing, or new product introductions. | | Supply Shaping | Strategies employed to influence the supply side of the equation to better match demand, such as adjusting production schedules, managing lead times, or optimizing logistics. |