Class 2

The case asks when should we pool and when does it matter?

“Where is the savings really coming from”

• Slow moving products don’t represent a lot of flow, so when you centralize, you save on inventory and spend on transpo, so you save on slow-moving products
• Measuring uncertainty: Standard deviation / mean
• A vs. B picture; what is my forecast error/avg. forecast.
• It’s all about the base on the XY

Measuring uncertainty mathematically:

Coefficient of Variation (CV) = σ/μ Where:

• σ is the standard deviation of demand
• μ is the mean demand

This ratio helps us understand relative variability across different products:

• High CV (>1): High uncertainty relative to mean demand
• Example: Fashion items where demand varies greatly compared to average sales
• Low CV (<1): Low uncertainty relative to mean demand
• Example: Staple products with steady, predictable demand

Common sense interpretation:

• If your forecast errors are large compared to your average demand (high CV), pooling becomes more valuable
• For products with steady demand (low CV), the benefits of pooling are less significant

Think of it this way: If you're typically selling 1000 units but could be off by ±500 (high CV), pooling helps manage this large uncertainty. If you're selling 1000 units and usually only off by ±50 (low CV), pooling offers less benefit.

Comparing Two Products with Different Coefficient of Variation:

In this comparison:

• Product A shows high variability relative to its mean (High CV)
• Greater uncertainty in demand
• Better candidate for pooling
• Product B shows low variability relative to its mean (Low CV)
• More predictable demand
• Less benefit from pooling

Adjacencies of Regions 4 and 5, and when low-demand products and regional demand (specifically, low demand) indicate that a region should be cut out and lumped into another region. Where uncertainty is large (so high SD and low mean)

When correlation coefficient increases (becomes more positive), pooling benefits decrease because:

• High correlation means both regions experience similar demand patterns simultaneously
• When both regions have high demand at the same time, a pooled facility needs to maintain high inventory levels
• Similarly, when both regions have low demand, the pooled facility has excess capacity

Pooling is most beneficial when demand patterns are negatively correlated because:

• When one region's demand increases while another decreases, they "cancel out" each other's variations
• This cancellation effect reduces the total variability the system needs to handle
• Similar to investment portfolio diversification - different stocks that move in opposite directions reduce overall portfolio risk

Mathematical explanation:

• With perfect positive correlation (ρ = 1), combined variance = (σ₁ + σ₂)², maximizing total variability
• With perfect negative correlation (ρ = -1), combined variance = (σ₁ - σ₂)², minimizing total variability

This explains why pooling is most effective when:

• Seasonal patterns are opposite (e.g., winter coats selling in different hemispheres)
• Economic cycles affect regions differently
• Product launches are staggered across regions

The cancellation of errors refers to how combining regions with different demand patterns can reduce overall variability. When one region experiences higher-than-expected demand while another experiences lower-than-expected demand, these deviations tend to offset each other, resulting in more stable overall demand.

• Example: If Region A is 100 units above forecast while Region B is 90 units below forecast, the combined error is only 10 units
• This effect is strongest when demand patterns are negatively correlated
• The cancellation effect reduces the need for safety stock in a pooled system

This principle is similar to portfolio diversification in finance, where combining assets with different risk patterns can reduce overall portfolio risk.

Longer lead times generally increase the benefits of pooling for several reasons:

• Longer lead times create more uncertainty in demand forecasting
• Greater uncertainty means higher safety stock requirements
• Pooling becomes more valuable as it helps mitigate the increased uncertainty

This occurs because:

• With longer lead times, the standard deviation of demand during lead time increases
• Pooling helps reduce this increased variability through risk aggregation
• The relative savings in safety stock become more significant with longer lead times

Therefore, companies with longer lead times (e.g., global supply chains, ocean freight) often benefit more from pooling strategies than those with short lead times.

Lead times: If your supply can supply instantly, then you have no forecast error, then there is no benefit of pooling, so the longer, the lead time increases variability, as forecast error gets bigger, then your safety stock increases. For local suppliers who are really fast. Think of components whose suppliers have really long lead times. The concentration of the supplier matters as well.

Target Cycle Service Level represents the probability of not stocking out during a replenishment cycle. It's a key metric in inventory management that measures:

• The probability of meeting demand during the lead time period
• The desired percentage of cycles that end with positive inventory
• The trade-off between service quality and inventory costs

For example, a 95% CSL means:

• 95% of replenishment cycles will not experience stockouts
• The business accepts a 5% chance of stockout during any given cycle

CSL directly impacts safety stock calculations:

• Higher CSL requires more safety stock
• Lower CSL reduces inventory costs but increases stockout risk

The relationship between CSL and safety stock is non-linear - small increases in CSL can require large increases in safety stock, especially at high service levels.

For Rolls royce and Ferrarir, they don’t carry at the dealer. You don’t carry a lot of inventory there, you make the car after people make the order. the bnefit of pooling there is minimal

More expensive product increases its holding cost.

As product cost increases, the benefit of pooling can be significant due to several factors:

• Higher holding costs mean greater financial impact from excess inventory
• Expensive products typically require higher safety stock investment
• Pooling can reduce total inventory investment while maintaining service levels

However, there are important considerations:

• Transportation costs must be weighed against inventory savings
• High-value products may have special handling or security requirements
• Customer expectations for expensive products often demand faster delivery times

Pooling reduces inventory, the higher the cost of inventory, pooling most valuable for expensive products. But there’s sometimes disconnect as if its expensive ai want to touch it and feel it.

• The Venn diagram is product + customer need —> channel/value prop
• Home Depot example of price sensitive vs. responsive/convenience sensitive
• Difference between fertilizer and wee whacker; For the convenience of not coming to the store, I’ll charge you $15 but for the weed whacker, I’ll then
• What’s the interaction between how the service delivery can change my perception of the product
• His three dimensions: W/ Price or service sensitive to identify the channel (convenience/return ability, etc.) `
• 1.) Product Value = (Value / weight) or (Value / Volume) → Will inventory holding cost be a bigger fraction of my cost or transportation. For fertilizer, transportation is higher % of the cost
• 2.) Demand uncertainty = (SD / Mean) Demand for fertilizer more predictable. Trimer has a lower base and less predictable.
• 3.) Information content of the product: When you tell me you wnat tactile, then you tell me there is info you can’t get from a description. High value products frequently correlate w/ high info content
• Ferrari, Hermes, etc. the centralized model is lower cost but they can get away w/ it b/c they have a prestige and brand story
• You don’t learn from the companies that succeeded, you learn more from the failures
• For diapers and detergent, the information content is low
• Costco decentralizes and sells fewer products at higher turns. Amazon, last mile shipping is not cheap.

What should we think of “unused capabilities” Fore example, high uncertainty and low value products are usually available to touch but you don’t care as much about their high information content

• Zales gets beat out by blue nile b/c it’s competing the wrong arena

"It's All About the Long Term" (1997) "We believe that a fundamental measure of our success will be the shareholder value we create over the long term."

This foundational quote established Amazon's commitment to long-term thinking over quarterly results, which proved crucial for their infrastructure investments.

"One thing I love about customers is that they are divinely discontent. Their expectations are never static – they go up." (2017)

This insight drives Amazon's continuous innovation and explains their relentless focus on customer experience improvements in their distribution network.

"In this turbulent global economy, our fundamental approach remains the same. Stay heads down, focused on the long term and obsessed with customers." (2008)

Written during the financial crisis, this demonstrates how Amazon's core principles guide their strategy even during market upheaval.

"There are two kinds of companies: those that work to raise prices and those that work to lower them. We will be the second." (2013)

This philosophy directly impacts their distribution strategy - constantly seeking efficiency to enable lower prices through scale and optimization.

"We want Prime to be such a good value, you'd be irresponsible not to be a member." (2016)

This shows how Amazon uses its distribution network as a competitive advantage, making Prime membership increasingly valuable through delivery speed and reliability.

Strategic Implications

• These quotes reveal consistent themes that shaped Amazon's distribution strategy:
• Prioritizing long-term infrastructure investment over short-term profits
• Using customer expectations to drive continuous improvement
• Leveraging scale and efficiency to lower prices
• Building sustainable competitive advantages through service quality