Managing supply and demand more effectively also requires better demand forecasting. The best way to achieve this is to segment products, channels, and customers according to factors such as volume, demand predictability, the product’s life-cycle stage, and supply chain impact. Since resources must be allocated across the product portfolio—and some areas will get more than others—this exercise can help companies prioritize their S&OP activities, investments, and attention. The result is a better balance between predictability and flexibility. It also makes it easier to predict the likelihood of stockouts or inventory pileups before they happen, allowing the organization to respond quickly. But to avoid needless complexity, segmenting must be a cross-functional effort instead of something that’s done differently within each function.
The first step is to identify, based on historical demand patterns, the products with highly predictable demand. Using statistical forecasting techniques, translate these predictable parts of the business into a demand plan. Then identify the products for which demand is consistently over- or underestimated. To address this type of “forecast bias,” speak directly to the salespeople who are most responsible for these errors and attempt to correct their behaviors. (See “The Need For Leadership,” below.) Finally, identify the products that sell in high volume but have very volatile demand. Work more closely with customers to improve the demand forecasts for these products.
An understanding of demand patterns can help improve the S&OP process by clarifying where management should direct its attention. For instance, products with high volume and high demand volatility might be reviewed monthly, while those with low volatility might be reviewed on a quarterly basis. When products are too fragmented to reveal any predictable demand patterns, channel-level segmentation may be the answer. Patterns may emerge when the fragmented products are consolidated by sales channel such as online, catalogue, or big-box store.
Segmenting products by demand predictability and volume can influence inventory management. For instance, products with highly predictable demand need less of a cushion and can be more tightly managed than products that are highly volatile. Further analysis might reveal that some highly volatile, low-volume products are minimally profitable and not particularly competitive in the market. Based on this cost-value analysis, a decision might be made to discontinue production and free up capacity for higher-value products.
Segmenting products can clarify the right strategy for each product type. A pharmaceutical company created product “buckets” on the basis of volume and demand predictability. (See Exhibit 3.) Each product segment was characterized by sales volume, revenue generated, and contribution margin. The analysis resulted in specific strategies for each product type—such as to maintain stability for high-volume products with low demand volatility, to closely track demand swings for high-volume products with high volatility, and to conduct deeper research on markets, competitors, and customers for new launches and high-volume products.
Segmentation can help determine how flexible production planning needs to be. If demand is predictable, for instance, a firm schedule can be developed. For volatile products, some flexibility in the schedule must be accommodated. Moreover, the insights from segmentation can help shape a set of rules or guidelines for managing the inevitable tradeoffs that occur when aligning supply and demand. Setting these guidelines in advance can minimize the need to escalate problems—and greatly speed up decision making.
Finally, segmentation can help companies set supply chain parameters based on the range of demand the system can absorb without breaking down. Consistently exceeding that level of demand can indicate a permanent new trend—and a need to reset the supply chain parameters by increasing production capacity, adjusting logistics, and so forth. Similarly, knowing which product supply chains can absorb incremental increases in volume and which cannot will help simplify S&OP planning.