Supply Chain Design and Planning PDF

Lecture 1 The supply chain is a (complex) network: - - - - - - - - Logistics costs are a big % of our final product price, and they keep rising due to longer supply chains, tight capacity and high demand SHEIN: - - - - The right supply chain strategy depends o...

Lecture 1 The supply chain is a (complex) network: - - - - - - - - Logistics costs are a big % of our final product price, and they keep rising due to longer supply chains, tight capacity and high demand SHEIN: - - - - The right supply chain strategy depends on the market you're in  Reshoring: - - Some questions of this course: - - By relocating suppliers closeby, you can increase efficiency regarding economies of scope Producing in a small scale, but remaining efficient Rising product returns are a big problem, especially in e-commerce - - - Huge increase in using AI for Supply Chain Management, like ChatGPT is used for everyday tasks Key goal for this course: what are the key elements that drive the solution to a problem  **Lecture 2** ### **How Push/Pull Postponement Strategy Works:** - - - Lecture 2 Supply chain integration strategies -\> make SC more effective Your input and output depends on your position inside a certain SC - Ikea: how does it operate its SC? - - - - - - - - - Tradational manufacturer: - - - - - - - - - In a push-based supply chain, the manufacturer bases orders on the orders from a retailer's warehouse  Appropriate supply chain strategies (push/pull) can change during a product's life cycle: - Difference between push-pull boundary and push-pull strategy: - - Identifying Push/Pull boundaries of certain companies, at what stage of the supply chain do companies switch from push to pull?:  3d printing: - - - - E-fulfillment: processing of online orders. the supply chain of the online retailer is vastly different, they suffer a way bigger return rate than normal stores Certain companies like Amazon (prime) use an extreme push strategy: they let you try their products by sending them to your home. They still make profit: when customers try them on they are more likely to buy it. - On the other hand, there are companies like Uber or Thuisbezorgd: they do not have a product. They use an extreme pull strategy: they only do something when a customer demand meets their supply. The Internet has a big impact on how companies like Walmart and Amazon sell their products - - - - Case discussion: the great inventory correction Crisis: too much inventory because of little demand for semiconductors Common problem: dropped sales and excessive inventory - Problem of altera: - - - - How has altera modified its strategy: - - - - - - - - - - - How do we calculate the total working processing time of the product? - - Disadvantages: Decreased customer satisfaction because of longer lead time Potentially longer lead time compared to customers being used to getting products right away Loss of economies of scale, we can\'t produce as a batch The software is only as good as the forecast Needs accurate input from customers, sometimes they order more than they need to make sure they have the priority of production Advantages: Lead time will decrease after the change during a stock-out The new strategy is more flexible Reduced planning cycle times because of i2 software A balance between efficiency and responsiveness Customers will react: - - - What information does flextronics have that its clients do not? How do they leverage this information? - - - - Case study The Great Inventory Correction discusses the severe downturn in the technology sector that began in late 2000, leading to a massive oversupply of inventory among high-tech Companies. This surplus resulted in significant financial write-downs for major companies The article highlights how the rapid decline in demand caught companies off guard, leading to a reevaluation of inventory and supply chain management practices. Companies started implementing strategies to make their supply chains more flexible and responsive, such as reducing inventory at different stages, improving collaboration with suppliers, and using supply chain management (SCM) software to better forecast and manage production cycles. Despite these efforts, the article underscores the challenges in accurately predicting demand in the volatile tech industry. For example: \- Not everyone have a supply chain system. \- Garbage in, garbage out. \- Some organizations will lie about their demand to seem to be more important The article concludes by noting that while the industry awaits the next upswing, companies like Dell and IBM, with strong supply chain management practices, are better positioned to extreme downturns in demand. Q1. How has Altera modified its strategy? Why? Classification: Internal Altera changed its strategy to avoid the large inventory losses it experienced by keeping products in a more flexible, unfinished form (die banks) and only completing them when customer orders are confirmed. The company also shifted to a build-to-order model for mature products, stopped producing new products on speculation, and searched for better visibility into customer demand. These changes were designed to reduce financial losses, increase flexibility, and improve forecasting accuracy. Q2. Do you think Altera's new strategy will be successful? What are some advantages and disadvantages of the new strategy? I think that Altera&\#39;s strategy will be successful. Advantages: It allows Altera to have a better adaption to the demand because customer orders are required. As well as improved collaboration with customers. Reduced inventory costs: there is less devaluating of inventory because altera will hold less stock. Meaning that if a new or mature product is not order of is being phased out Altera will not be left with obsolete stock. Disadvantages: Longer waiting times for new products. As the intermediate products still have to be made into finished products after the customer has ordered them. Risk of lost sales: when demand rises quickly Altera may not be able to match the demand and risk the customers looking for the product elsewhere. Q3. How do you anticipate Altera's customers will react to this strategy? What are the advantages and disadvantages for Altera's customers? I think the customers will have a mixed feeling with this strategy Altera applies. An advantage of this strategy for customers is that they can order a product that will be build specifically to their demand and not a standard PLD. Another advantage is that the chips will possibly be cheaper, because the lead time and stock are decreased. On the other hand, there will be longer lead times because the products are built to order. This is a big disadvantage for customers who need the products immediately. Classification: Internal Q4. What information does Flextronics have that its clients do not? Why? How can Flextronics leverage this information? Flextronics has specific information about demand patterns, and trends in the market. The main reason for this is that they are a very big company, with a wide range of products (from printed service boards to a cellphone) and a range of clients (Cisco, Lucent, Nortel, Ericsson). This gives them a unique information position, which gives them an information advantage. Q5. How does IBM manage its suppliers in order to make its pull strategy more effective? IBM provides its suppliers with visibility into its inventory levels. This transparency enables suppliers to respond more quickly to IBM&\#39;s needs, supporting the company&\#39;s just-in-time (JIT) production approach. Additionally, IBM has reduced its stock keeping units. The same parts are used in different products, with this IBM simplify its inventory management. This leads to quicker assembly and production, making the pull strategy more responsive and efficient. IBM also keeps the number of suppliers small. Purchasing is structured by category, with a market expert assigned full-time to each category. This arrangement ensures a continuous focus on all parts and suppliers. IBM has a very detailed forecast for the next 90 days, a fairly detailed forecast for 90 days to a year and a strategic forecast for longer periods. These forecasts enable IBM to prepare for upcoming demand, making the company&\#39;s pull strategy more effective. How much to produce depends on the demand forecast Push based supply chain - - - Pull based supply chain - - -  Economies of scale are cost advantages realized by companies when production becomes more efficient. A screenshot of a computer Description automatically generated  Amazon pushes the products to the distributors (Push), but only delivers when there is an order (pull) A diagram of a production process Description automatically generated Mcdonalds its push pull boundary is at production. Mcdonalds will start production when there is an order Nasa will start designing when there is money/an order Push pull boundary is the same as the klant order ontkoppel punt in dutch (KOOP) A diagram of a diagram Description automatically generated  Three companies on the left have a push strategy Three companies on the right have a pull strategy **Lecture 3** Supply chain VS development chain: - - - Product introduction frequency: in relation with technology clock speed - We will use integral product architecture: - Focus on the value and performance quality of processes used to deliver goods and services: - - - - - - - - - - - - - Transforming the process: Traditional manufacturing: - Concurrent/parallel processing: - Design for logistics: - - - Design for environment: - - - - Trade-offs: Design for manufacture: - Standardization: - Design for logistics: - Design for environment: - Concurrent and parallel process: - Before, suppliers have been selected after design of product or components But now: firms find benefits in involving suppliers in the design process Different strategies we can use:  - - - Dell: black box, asus designs it for them P&G: between white and grey, more complex products are designed together with suppliers Toyota: grey box, dashboards are designed together with suppliers Making the relationship a success: - - - Which suppliers an be integrated? **Case discussion 2:** Global printer market: highly competitive, but HP has the biggest market share - - - - - A printer is an innovative and expensive product - - - - They try to reduce the amount of customization they have to do, in order to postpone this customization to the last step The power supply issue: - - - - - - - - The supply chain was changed in aspects: - - The push/pull boundary did not change, everything is produced based on forecasted demand. It is still a push-based supply chain Delayed product differentiation: - - - - - - - - - - - - In the ramp-up stage, a loss of sale is very high, we should have a high service level that covers the majority of the demand - - - Aggregate the total benefits with the costs, or adopt a hybrid strategy: when the benefits outweigh the costs, we do it and otherwise not Hybrid strategy: - - What hp did: - - - - - **Lecture 4** Smart pricing: how to set up prices for different types of products, and design a pricing strategy to get more revenue from customers Wendy's price surging (discount) strategy: spread out the demand, during peak hours it will be less busy because of higher prices, and customers will want to buy the product for a lower price at quiet hours Using price optimization strategies When p = 1600, D = 200 Revenue: when p = 1000, R = 500.000 How to maximize revenue: R = p \* D\ R = DP = 1000P - 0.5p\^2\ Take derivative:\ dR/dp: 1000 - p = 0\ p\* = 1000, R\* = 500.000 Demand is not deterministic: if we make a forecast error and have more inventory we can sell at the set price, we will lose a lot of money - - - - - Markdown = promotion/discount - Revenue management: charging different prices to different customer groups to maximize the revenue Business and first class were developed to distribute the revenue for different types of customers so that the total revenue will increase: - 1. a. 2. b. 3. c. 4. d. 5. e. f. 6. g. h.  WILL COME ON THE EXAM 1. a. b. c. 2. Price optimization can be extremely complex: - - - Price optimization can also be very risky - Rue Lala has problems with setting prices - - - - Challenge 1: estimating lost sales due to stockouts Solution: - - - - Challenge 2: predict demand for items that had no historical sales data Solution: - - -  Dynamic pricing What if we allow price changes during an event? - - - - - - - - - To increase benefit of dynamic pricing: - - - - - - Uber has the ability to adjust the supply, but Wendy's does not: - - - - **Case Discussion 3:** The great rebate runaround Rebates: a cashback customers receive after a certain period Why rebates are an example of customized pricing (differential pricing strategies): We distinguish three groups: - - - To make sure people don't switch from yellow to green: make complex rules so that people in the yellow group will not find it too easy to apply for a rebate 40% of all rebates never get redeemed - - - If not all rebates are redeemed, you retain a part of the initial price that people never cash in, which you lose anyway if you provide a discount upfront Even if all rebates are redeemed, it is still better than decreasing wholesale price: - - - But what if customer demand is smaller than the order quantity of retailers? - - Companies eliminate rebates, because: - - - - - New strategies: - - - - - - **Lecture 5** Circular economy: -  Challenges in the reverse supply chain: - - Companies are moving from products to servitization - - - - - - - - - - Benefits for customers: - - - - A lot of companies when they first started - - - They adopted servitization - - - - - Companies expect a linear increase when they invest - - Due to competition, there are a lot of shared bike companies - - - - Uber has increased carbon emissions than before it was introduced **Rebound effect**: when something is cheap and accessible, there will be an increased demand for the number of vehicles that are actually needed - **Science-based carbon targets**: when firms realize they need to reduce carbon emissions - - - - - Problem: people are looking at the correlation but not at the causation - Differences between firms with and without science based targets: - - To be in line with SBTi criteria, companies must set scope 3 (carbon emissions from up-and downstream supply chain) targets - Set up targets for different categories - - Name examples of why sustainability projects fail -\> FOR THE EXAM Over-consumption: - - - - - - **Case 5 discussion** Anderson was the CEO of Interface, a carpeting business (one of the most polluting at the time), and he introduced a strategy to drastically reduce the pollution of his company - - - Should interface move into services? Are they ready for it? - - - - - - - - Evergreen services agreement: close the loop by preventing used carpet materials from ending up in landfills - - - - - Why did the negotiations with the university of texas at houston break down? - - Why is interface finding it challenging to sell Evergreen Services Agreements? - - - - - - - What do you think Interface should do? - - - - - Why a product-service bundle with leasing does not work for carpets? What industries can benefit more from this strategy? - - - **Lecture 6** **Sales and operations planning**: - - - - - - - - Overview of major operations and supply planning activities:  **Aggregate operations plan**: translates annual and quarterly business plans into broad labor and output plans for the **intermediate-term** - - - - - - - - Production planning strategies: - - - - - - - - - - - - - - - - - - Required **inputs to the production planning system**: External to firm (can impact us from the outside): - - - - - Internal to firm (we can control): - - - - Apple cuts back production of Apple Vision Pro because of a too-complex design Sales and Operations planning processes and performance measurements: S&OP process: 1. 2. 3. 4. 5. S&OP is cross-functional work: how do we align the interests Cross-functional teams - - - S&OP meetings - - - What can go wrong: - - - - - - - - - How to overcome these challenges: - - - Future trends in S&OP: - - - - - - - **Case discussion lecture 6** Leitax experienced forecasting errors because they had no formalized S&OP department Q1: Which individuals or functions should be held responsible for the crisis in 2002?  Every department created their own forecast, chasing their own incentives Top management can also be blamed, for not aligning the different departments Q2. How does the new forecasting process work? - - - - - - Overall benefit of this new process: - - - - - - So, it was better in terms of: Accuracy: - - - - Alignment: - - - - Support for other planning processes like financial planning, manufacturing scheduling, and capacity planning Q3: what happened with SF6000 and ShootXL/Optix-R? What happened that lead to mistake in forecasting? - - - - - How do you assess giving more influence to statistical forecast or the DMS function to get more accurate forecasts? - - - - So, we can still make it more accurate with statistical forecasts, but it will not be adopted by them because they dont trust it: we need alignment What about other possible changes to the process, for example using life cycle forecasts? How do we evaluate these options? - - - **Lecture 7 - Guest Lecture** Eyeon - Forecast & Demand Management Ultra Fresh: manufacturer and supplier of fresh meals - 5 factories for the pre-processing of vegetables, processing and packaging of finished products - The customers of Ultra Fresh are the big supermarket chains in the Netherlands - Product portfolio of 1500 unique products and the planning is currently still done in Excel - The company wants to improve service levels to the customers and asked for support to setup demand planning in the organization Focus on **Demand management** Demand forecasting: looking at patterns in history and estimate whether it will continue in the same way in the future - - The goal of demand management: 1. - - 2. 3. Importance of demand planning: Short term: determine what to produce when Mid term: how much vegetables and package materials to buy Long term: strategic decisions Different decisions per term Best-in-class demand management covers 4 dimensions: - - - - **Step 1: Create a synchronized demand planning cycle** - - - - - - - FACTS (review historical data + forecast projection) -\> EXPECTATIONS (enricht forecast) -\> EVALUATE The shorter the lead time, the more accurate a forecast will be The process is a cycle: you create a forecast, plan a demand meeting, and evaluate the results, after a certain period, the cycle repeats Applying Segmentation: increasing your effectiveness by segmenting all of the products, so you know what to focus on - - Measure & evaluate forecast performance - - - **Step 2: implement a fit-for-purpose demand planning solution** Key challenges in planning environment: heavy workload, lack of E2E visibility, lack of scenario planning, data latency, long implement cycles, IT dependent Planning tools: one size does not fit all. Compare business size with business maturity Criteria for the selection of a planning solution: - - - - - - Product life cycle: demand planners should take end-of-life products into account (stop selling them)  NPI: New product introduction ABC: Segmentation of products, product portfolio EOL: End of life XYZ: type of demand -\> how stable or volatile is your demand **Step 3: let demand planners take control of their role to orchestrate the Demand Planning cycle** The role of the demand planner will change due to new technologies, their work is changing and their added value is increasing Demand planners should effectively analyze forecasts, and be a change agent: they should work together with different departments - - - - - - Human behavior is all throughout the demand planning cycle Almost all adjustments are small and upward, and these do not have a significant effect The downward adjustments often do improve the forecast **Step 4: apply advanced statistical forecasting techniques to improve accuracy** **Time series models (basic)**: with stable demand and patterned volatile demand **Driver-based forecasting (advanced)**: with volatile demand that cannot be predicted by historical data, this forecasting technique takes other factors into account - - - - Ultra Fresh can make this step to driver-based forecasting, but they should first grow in their maturity **Lecture 8** Key inventory decisions: - - Why is managing inventory correctly important? - - Why is inventory required? - - - - - - - - - - - - **Economy lot size model** - - - **= Optimal order** **Demand uncertainty**: - - - - - - - - - Single period models (**newsvendor model**) - - - - -  How much to order to maximize expected profit: Above, below or equal to the expected demand -\> go for the optimal order quantity which maximizes expected profit **Calculate average expected profit given a specific order quantity** Compare **marginal profit** of a sale and **marginal cost** of not selling - - - - - Continuous demand:  Example:\ Assume there is a normally distributed demand with a mean of 90 papers, and a standard deviation of 10.\ How many papers should we order if we want an 80% chance of not stocking out Answer: We should look in the corresponding z table for normal distributions, we find that for the 80th percentile, there is a z-score of about 0.85 This means that we need approximately 0.85 standard deviations of extra papers to have an 80% chance of not stocking out So, we need 0.85 \* 10 = 8,5 -\> 9 extra papers\ We order a total of 99 papers to have an 80% chance of not stocking out To find the optimal probability that an additional unit will not be sold (maximized order to requested demand) -\> cost of understocking/(cost of overstocking + cost of understocking) This ratio is the order that minimizes the expected mismatch costs, which in turn maximizes profit, in a concrete formula it looks like this: Q\*(optimal order quantity) = mean of the demand + z of critical ratio \* standard deviation of demand ------------------ Q\* = μ + z \* σ ------------------ Now, you minimize mismatch costs, and maximize the profit at the same time **Risk-Reward Tradeoffs**: - - - - - What if the manufacturer has an initial inventory? - - - - - **Multiple Order Opportunities - Continuous review policy** - - - -  How to determine R and Q? - - - Order size (order quantity) is driven by balancing the inventory holding costs and fixed order costs, order point depends on lead time uncertainty **Multiple Order Opportunities - Periodic Review Policy** - - - - - A: short intervals (e.g daily): **(s, S) Policy** - - B: intervals (e.g weekly or monthly): base-stock level policy - - - - -  Base stock should be enough to cover demand variability for as long as your lead time is (until your next order arrives) **What is the appropriate level of service?** - - - - - - - - - - - **Risk pooling: when we aggregate demands the total variability is low** By combining or \"pooling\" demand from **multiple sources or locations**, a company can *reduce the overall uncertainty or fluctuation in demand*, making it easier to manage inventory efficiently - - - - - - - - - - - - Other forms of risk pooling: - - - - - - - - - - - - - **Case discussion lecture 8** **Obermeyer sport** Problem: Demand is uncertain, and lead times are long because of seasonal products Clothes are made by a joint venture in China and Hong Kong - Q1: Using the sample data given in Table 2-20 (and posted on Canvas) make a recommendation for how many units of each style Wally should make during the initial phase of production (assume all 10 styles are made in Hong Kong and ignore price differences among styles) Answers: - - - - - - - - Q2: what operational changes can Wally implement in order to improve performance? Answer: - - - - - - Decrease minimal order quantities: to reduce mismatch costs and reduce risks **Lecture 9** **supply contracts** - - - Make-to-order: Wholesale price contract: price at which the manufacturer sells the product, and the retailer decides how much they order Retailer: places order depending on forecast Manufacturer: receives order and satisfies it Selling price, wholesale price, salvage price  Retailer will look at the expected profit in certain situations, and places the Q that maximizes profit Supply chain profit: sum of the retailer and manufacturer's profit Why is the supply chain profit maximized at 16.000, but the retailers profit at 12.000? The more the retailer orders, the more they sell, but the risk of not selling will increase, so eventually profit decreases To maximize profit, we look at the costs and profit margin (understocking and overstocking) 86% probability of satisfying all customer demand -\> maximum profit for the supply chain **Double marginalization**: Can we set up better contracts to achieve the optimal service level of 86% - **Buy back contract**: Supplier agrees to buy back unsold goods from the buyer for some agreed-upon price Implications: - - **The buyer buys more because their risk is reduced, so the supplier might be able to make more profit** The cost of overstocking is lower -\> the order quantity is increased Swimsuit example: the manufacturer offers to buy unsold swimsuits from the retailer for \$55 New expected profit after the buy-back contract:  The profit is also larger for the manufacturer, he takes risks but it pays off profit-wise  **If the buy back price increases to 65, the retailer will order 16.000 units, so we achieve the maximum profit of the whole supply chain** **Revenue sharing contract**: - - - - - - - -  Both expected profits increase after the revenue sharing contract -\> they share the risk  Implementation drawbacks of supply contracts: - - - - - Now the retailer orders more, thats better for both parties and for the entire supply chain - Other types of contracts: - - - - - - - - Now we are in a make-to-stock setting Manufacturer produces before they receive the final order - Make-to-stock setting: - - - - Distributor cost: p (selling price) = \$ 125 per unit w (wholesale price) = \$80 per unit Manufacturer cost: K = \$100,000 (fixed cost) c = \$55 per unit (variable cost) s (salvage)= \$20 per unit Demand forecast:  Manufacturer maximum at 12.000 units Distributor expected sales increase on higher capacity of the manufacturer Supply chain optimum is at 14.000 units Service levels: Optimal Service Level: CSL\* = Cu / (Cu+ Co ) Manufacturer: CSL\* = (80-55)/(80-55+55-20) = 42% Distributor CSL\* = 100% Supply Chain: CSL\* = (125-55)/ (125-55+55-20) = 67% Which service level will be offered? - - **Pay-back contract**: the buyer agrees to pay some agreed-upon price for the inventory produced (capacity set) by the supplier but not purchased (used) - - - Ski-Jackets example: The distributor offers to pay-back the manufacturer for each unsold ski-jacket \$18.  - **Cost-sharing contract:** the buyer shares some of the production cost with the supplier, in return for a discount on the wholesale price - - Ski-Jackets example: the distributor pays 33% of variable production costs; the wholesale price drops from \$80 to \$62  Implementation issues: - - - - - Credible information sharing is important: - - - - - Contracts with asymmetric information: **Capacity Reservation Contract** - - - **Advance Purchase Contract** - - - - - **Contracts for Non-Strategic Components** - - - - - - - - - - - - **Portfolio of contracts:** - - - - - - -  **Case discussion lecture 9** **American tool works**: they produce handtools, and sell to dealers (80%), and direct customers - - - - - - - - - - - What drives sales (based on channel) - - - - - - - **Question 1:** what can ATW do to increase inventory and sales at its small and medium-sized dealers? - - - - - - - - - **Question 2:** should ATW adopt similar approaches (like offer sales people incentives) as its competitors and what other approaches should they consider to manage their distribution channels? - - **Lecture 10, the value of information** Bullwhip effect: - - Operational causes of the bullwhip effect: 1\. **Demand Forecasting** - - - 2\. **Lead Time** - - 3\. **Batch Ordering** - - 4\. **Price Fluctuations** - - 5\. **Order gaming** - - **If all operational causes are removed → the effect persists** The bullwhip effect persists in an environment with *stationary* and *known* demand - - -  Bullwhip: Behavioral Causes 1\. Overreaction to backlogs (receiving orders you can't satisfy) - - 2\. Decision-makers under-weight the supply line - 3\. Lack of trust - 4\. Bounded rationality - - Key insights bullwhip effect: 1. 2. 3. 4. 5. 6. 7. **Quantifying the Bullwhip** Retailer follows a simple periodic review policy (base stock policy) - - - - - - - - When p is large and L is small, the bullwhip effect is negligible (the ratio is low). Effect is magnified as we increase the lead time and decrease p.  If the lead time is low, the ratio is higher (but now so much for small values of p) If the forecasts are less stable (even though it might be desirable), p is small -\> the longer the lead time -\> the longer the impact on order variability KNOW WHAT THE FORMULAS MEAN (NOT MEMORIZE IT) The Impact of Centralized Information **Centralized demand information -\> everyone has the same information about customer demand** - Same ratio, but calculating it for specific stage (k) in the supply chain As k increases, the further away you are away from the customer -\> the sum of the lead times between the customer and each stage of the supply chain **Decentralized demand information -\> information is not known** -  The effect is much worse, because the effect of the lead time is higher *Centralizing demand information* can significantly **reduce the bullwhip effect** - In the first stage of the supply chain (k=1) your order variability is lower compared to fifth member of the supply chain(k=5), and it shows that centralized information lowers the bullwhip effect and order variability significantly - Methods for **coping** with the bullwhip (to contain the variability) - - - - - - - - - - - - - - - - Information for **Effective Forecasts** **Information also can have a direct effect on forecasting accuracy** - **Pricing, promotion, new products** - - - **Collaborative Forecasting addresses these issues (e.g. CPFR)** - - Information for Lead-Time Reduction - - - - - - - - How? - -  - - - Shorten Lead Time: **Quick Response** - - - - - - - - - Locating Desired Products - - - - - - - Internet of things (IoT) - - - - - - - - - - Blockchain technology (to increase visibility in our supply chains): - - - - - - - - - - - - - - - Decreasing marginal value of information - - - - - - - - - - - - **Case study Reebok NFL replica jerseys** - How should Reebok plan and manage inventory to manage costs while providing the flexibility required to meet demand for NFL replica jerseys? Reebok: - - - - - - - Purchasing cycle: - - - Q1: given the uncertainty associated with player demand, how should reebok approach inventory planning problem for NFL replica jerseys? - - - - - - - - - Q2: What should Reebok's goal be? Should Reebok minimize inventory at end of season? Or maximize profits? Can Reebok achieve both? What service level should Reebok provide to its customers? - - - - - - - - - - - Q3: Are the models in Section 2.2.2. helpful here? What is the cost of underage for a dressed jersey? What is the cost of overage for a dressed jersey? How might Reebok decide between dressed jerseys and blank jerseys? Q4: Using the forecast for the New England Patriots -- what is the optimal quantity to order for each player? For blank jerseys? What profit do you expect for Reebok? How much and what type of inventory is expected to be left over at the end of the season? What service level? - - - - - - - - - - **Lecture 11, distribution strategies** Two options: 1. a. i. ii. iii. iv. b. v. 1. vi. 2. c. vii. viii. d. ix. x. xi. e. xii. xiii. Central vs local facilities Centralized facilities: - - Factors for location decision: - - - - - - - - - - Issues with cross-docking (directly transferring goods to distributors with no storage in between): - - - - - - - The impact of uncertainty aggregation 20 stores (5 per DC) SL: 97.5% Store Demand = 50+- 35 (std) Replenishment time = 1 week Each DC Demand = 5 50 SQRT(5) 35 = 250+-78 CDC Demand = 20 50 SQRT(20) 35 = 1,000+-156 Total stock at stores = 20 (50+1.96 35) = 2,372 (How much inventory do we need in total in all 20 stores to satisfy 97.5% of demand) Total safety stock at DCs = 4 (250+1.96 78) = 1,612 (how much inventory we need if we store everything in the distribution centers) Total safety stock at CDC = 1,000+1.96 156 = 1,306 (how much inventory we need if we store everything in 1 big central distribution center This shows that when we move inventory more centrally, we need less safety stock -\> as we aggregate demand, the std of the demand does not increase linearly) -  - - - - - - - **Centralized pooled systems perform better:** For the same inventory level, a centralized system provides: - - Push-pull supply chain - - - End consumers will see better customer service -  - - - Customer search - - - Impact on inventories: - - - - - - - - - - - - -  - - - Effect of α (=% of customers searching the system) on orders  As α (customer search) increases **Each retailer's order and profit increases** **Retailers**' expected profit: **Always higher in the centralized pooling system** than in the decentralized system **Manufacturers**' total expected profit: For α \> α\*: higher in the decentralized system For α \< α\*: higher in the centralized system **Manufacturer always prefers a higher search level!** - - - - - - **Transshipment**: Instead of customer going around, the inventory goes around - - - - - - **When we have retailers with different owners**: By setting good transshipment prices, we can still achieve first best optimal solution - Ownership structure is important: loss of control - - - - - - - - **Lecture 12 - guest lecture supply chain resilience** Boston Consulting group: specialized in consulting in business topics regarding artificial intelligence This guest lecture: How resilience is being designed and implemented in companies External environment: - - - - What is supply chain resilience - - - - - - - - - - - How can an organization build supply chain resilience - - - - Multiple sourcing: having multiple suppliers as a back-up plan in the case one of them falls out Key pillars to build supply chain resiliency:\ sensing: 1. 2. 3. adapting: 4. 5. 6. thriving: 7. 8. 9.