Unlock Your supply chain Potential

Managing intermittent demand, often referred to as lumpy demand, presents significant challenges for supply chains. Unlike the steady, predictable patterns of fast-moving items, intermittent demand is characterized by irregular spikes and frequent periods of no demand at all. Traditional forecasting and inventory management methods struggle to provide the accuracy and stability needed to manage these unpredictable items, which are often found at the "long tail" of a company’s product portfolio.

Why Traditional Methods Fall Short

Conventional inventory management techniques—like time-series forecasting or ABC classification—are based on assumptions that demand follows a normal distribution. These methods work well for fast-moving items but fall apart when applied to slow-moving, lumpy items. They tend to either overstock or understock products, leading to excess inventory in some areas and stockouts in others, all while failing to meet service levels.

This happens because traditional solutions rely on aggregate data and broad assumptions that don’t reflect the real-world variability of intermittent demand. The result is misaligned inventories, with supply chain managers constantly making manual interventions to compensate, often leading to inefficient inventory levels.

Understanding the Causes of Lumpy Demand

Several factors contribute to the growth of lumpy demand:

1. Product Proliferation
   The introduction of more product variants has significantly increased the number of SKUs, but not necessarily overall sales. This dilutes demand across a broader range of products, leading to higher variability for individual items.
2. Shorter Replenishment Cycles
   Frequent deliveries and more granular forecasting create shorter time buckets for demand, increasing the likelihood of variability. An item may appear stable on a monthly level but behave unpredictably when viewed on a daily or weekly basis.
3. Focus on Total Supply Chain
   Increased collaboration across the supply chain has led to disaggregated demand streams. With demand being forecasted at more granular levels, such as for specific retail locations or distribution centers, variability increases.

How to Tackle Intermittent Demand

The key to handling lumpy demand lies in moving beyond deterministic models that assume normal demand patterns. Instead, probabilistic solutions are better equipped to handle this variability. Here’s how to tackle intermittent demand effectively:

1. Use Probabilistic Demand Models
   Probabilistic models analyze the full range of demand behavior, including variability and skewed distributions. These models are tailored to each SKU-location, capturing the unique demand patterns rather than applying a one-size-fits-all approach.
2. Inventory Optimization with Service Level Focus
   Traditional methods like ABC classification are not sufficient for intermittent demand. Instead, companies need dual-engine demand and inventory models that synchronize inventory levels with demand patterns. This helps ensure that both fast and slow movers receive the appropriate attention, minimizing excess stock while maintaining service levels.
3. Leverage Advanced Analytics
   Advanced inventory models take into account demand variability and frequency distributions. These models avoid the gross approximations used in traditional systems, providing a more accurate understanding of stock needs and enabling better decision-making.
4. Dynamic Replenishment
   For slow-moving items, using dynamic replenishment strategies is critical. Probabilistic solutions can model demand variability and recommend inventory levels that reflect real-world behavior, reducing the need for manual interventions and eliminating the bullwhip effect.

The Bottom Line: Adopting a No-Shortcuts Approach

Tackling intermittent demand requires a no-shortcuts approach that fully understands the complexities of demand behavior. Traditional models that rely on normal distributions and aggregate data are not capable of handling the variability inherent in lumpy demand. By mastering demand and inventory variability, companies can achieve higher service levels and reduce stock imbalances.

In the end, probabilistic solutions offer the best path forward for managing lumpy demand. These solutions provide the detailed, SKU-level insights needed to optimize inventory and ensure that companies meet their service level goals, even for long-tail items.