### Final Grade & Feedback Q1: 15/15 Q2: 15/15 Q3: 15/15 Q4: 10/15 [No quantitative example provided] Q5: 10/10 Bonus: 0/10 **Total: 65/80** --- 15.778 OPERATIONS MANAGEMENT Case 2 Yedioth Team Penguins Section B Dmitry Shikhov Joe Suh Michael Ben Aharon Sandra Liu Shota Furukawa 1.​ In the current distribution model, where each retailer is supplied once (a week) independently of all other retailers, what would be a good method to compute the quantity shipped to each retailer to guarantee that 99% of customers will be served? Apply your approach to compute recommended quantities to the 50 retailers (explain the methodology in the body of the report and provide the results in appendix). As magazines are distributed only once a week and not replenished within a week, each retailer has to manage their own stock level. Therefore, the stocks of magazines at each retailer must be computed to ensure sufficient stock levels with 99% confidence. Chart 1-1 As shown in Chart 1-1, each retailer has different expected sales with variance. We assume that the distribution of sales across retailers conforms to a normal distribution. Also, given that the dataset is uncensored, we assume that sales in the dataset reflect both realized and unrealized demand, including lost sales opportunities. Accordingly, Sell-through is considered to be negligible for the purpose of this analysis. We compute 1 each q* (average sales + safety stock) with the following steps and calculations detailed in Chart 1-2. Firstly, we compute the Average Sales μ and the Standard deviation σ of each retailer. Secondly, we compute Z based on the required confidence level, which is 99% in this case: Z is 2.33. q* can be calculated by q* =μ + Z⋅σ. As q* is the stock for retailers, q* should be an integer and should be rounded up. Lastly, we sum up the q* of all retailers, which is the safety stocks for the entirety of Yedioth's magazine business. Therefore, the required stock (q*) should be 419 magazines per week. 🚨correct ~419🚨 Chart 1-2 2 2.​ If Yedioth could implement full pooling among all of the 50 retailers what would be the estimated benefit in terms of total production levels and returns (assume that the required service level is 99%). Note: Full pooling means that somehow all of the retailers could be supplied in real time from the same pool of inventory. In full pooling, we can consider the whole process as one single stock pool and ignore the variability of each retailer. The required stock (q*) should be determined to cover the variability of the total number of magazines as a whole system. As shown in Chart 2-1, the total number of magazines sold per week follows a normal distribution, and the same calculation as Q1 is applicable. As shown in Chart 2-2, the required stock q* = μ + Z⋅σ, which is equivalent to 237 magazines per week. 🚨correct pooling🚨 Yedioth should produce this amount to cover the variance, and the estimated returns will be 47.4 magazines per week, given the expected sales amount is 189.6 magazines per week. Compared to the retailer model in Q1, the production levels reduction and the returns reduction are expected to be less, as follows (Chart 2-3): ●​ Production levels reduction: 419 - 237 = 182 magazines per week ●​ Returns reduction: 214.4 - 47.4 = 167 magazines per week Chart 2-1 Chart 2-2 3 Expected Sales (1) Mean (μ) Expected Return (1) Safety Stock (1) + (2) Production (q*) Original 189.6 102.8 292 Retailer 204.6 214.4 419 Full Pool 189.6 47.4 237 Sales Agent* 189.6 103.4 293 Chart 2-3 3.​ Suppose that one could implement full pooling only among retailers that are treated by the same sales agent, what would be the potential benefit in terms of production levels and returns (assume 99% service level). Compare to 2) above. This solution should be the middle ground between the distribution models in Q1 and Q2 because Yedioth will stock up magazines with the sales agents model, between the full pool and retailers. As with the Q1 computation, it is required to compute the required stocks with a 99% confidence level for each sales agent. Firstly, we need to aggregate the sales by sales agent, and then compute the Average Sales μ, the Standard deviation σ, and the required stock q* of each sales agent. As in Q1, q* should also be rounded up, as partial magazines cannot be sold. (Shown in chart 3-1) As a result, the required stock q* should be 293 magazines per week. 🚨acceptable agent pooling🚨 The impact of production and expected returns compared to the retailer model in Q1 can be computed as follows. ●​ Production levels reduction: 419 - 293 = 126 magazines per week ●​ Returns reduction: 214.4 - 103.4 = 111 magazines per week As calculated in Q2, the benefits of the Sales agent model are below. ●​ Production levels reduction: 419 - 237 = 182 magazines per week ●​ Returns reduction: 214.4 - 47.4 = 167 magazines per week Compared to the full pooling model in Q2, the benefits diminished, while it improves the production efficiency and reduces returns. 4 Chart 3-1a 4.​ Propose more realistic policies that leverage the fact that the sales agent visits each retailer in the middle of the week. What would the benefit be of these policies? Leveraging the sales agents' mid-week visits for replenishment is a realistic and effective strategy for Yedioth as it can significantly reduce required inventory levels and improve overall operational efficiency over Yedioth’s current supply of each retailer only once at the beginning of the week. Since sales agents already visit retailers mid-week, we can use this visit to deliver additional inventory: one forecasted at the start of the week, and another based on updated sales information gathered mid-week. Replenishing mid-week helps absorb sales variability and enables a reduction in safety stock as a whole. Hence, this can also reduce the return rate (currently at 36%). Realistically, Yedioth can apply the stock pooling at the sales agent level (system in Q3) and ask them to replenish in the middle of the week. 🚨specific mechanisms🚨 This reduces the uncertainty in each forecast because the standard deviation of demand increases with the length of the forecast period, with shorter forecast intervals leading to lower demand variability. The Newsvendor model accounts for this variability when determining optimal stock levels, so less variability means we need less safety stock to achieve the same 99% service level. This approach also delivers several additional benefits: ●​ Lower return rates: Fewer unsold magazines need collection and scrapping, reducing waste and reverse logistics costs.​ ●​ Lower production levels: Less safety stock is required upfront, so printing large buffer quantities to cover weekly uncertainty is no longer necessary.​ ●​ Improved data quality: Mid-week visits generate uncensored demand data, improving forecasts and enabling better initial shipment planning in future weeks.​ 5 ●​ Revenue protection: Topping up stock mid-week minimizes stock-outs, protecting advertising-driven revenue and aligning with sales agents’ incentives.​ ●​ Retailer relationships: Proactive replenishment strengthens partnerships with retailers by ensuring consistent availability and responsiveness. Moreover, these improvements free up printing capacity for other business and revenue-generating opportunities. We can enhance this approach further by combining it with sales agent-level pooling. Pooling reduces variability across retailers within each agent’s group, and mid-week replenishment allows us to refine stock decisions using real-time data. These strategies create a responsive, efficient system leveraging existing infrastructure without the need to adopt the greater complexity of full central pooling. 5.​ What do you think are the organizational challenges that Assaf will have to address? Assaf faces significant organizational challenges 🚨multiple stakeholders🚨 as he works to implement changes that go against Yedioth's long-standing practices and culture of overproduction. A key obstacle is overcoming resistance from sales agents and retailers. Sales agents are currently compensated based on sales volume, which creates a strong incentive to oversupply in order to ensure every potential sale is covered. For both agents and retailers, the perceived cost of understocking is much higher than the cost of overstocking, so Assaf must revisit the salary and incentive structures to encourage minimizing returns and accepting occasional opportunity loss in the case of unexpected demand spikes. This could include redesigning compensation schemes to better balance the costs of understocking and overstocking. Since Yedioth provides full refunds for unsold inventory and bears the cost of scrapping, retailers and agents currently have little financial risk or “skin in the game,” reinforcing their tendency to over-order. Assaf should promote mid-week replenishment as a realistic and efficient solution. Adjusting inventory mid-week based on updated sales information allows Yedioth to reduce safety stock, decrease total production, and lower return volumes. Implementing technologies such as EDI for larger retailers and RFID or smart stands for smaller ones can support this transition by providing timely sales data and improving forecast accuracy. However, for these tools to have impact, sales agents and retailers must be empowered not just to report demand but to actively adjust inventory levels in response. In parallel, Assaf should reassess the current 99% service level target. Because the relationship between service level and safety stock is non-linear, even a small reduction in the target service level can yield a disproportionately large drop in required safety stock and overall costs. Combined with changes to incentive structures, optimizing the service level can help align operational targets with the true cost trade-offs of understocking versus overstocking. 6 Another organizational challenge lies with the Research Department, which sets shipment quantities. Their bias toward overproduction is driven by fear of lost sales and the difficulty of forecasting demand accurately, especially when true demand is hidden by censored data during frequent stockouts. This instinct to oversupply is reinforced because sell-through events often mask unmet demand, prompting them to inflate quantities as a precaution. When comparing models, the full pooling scenario stands out as the optimal target. It maintains the same expected sales as the current system but achieves them with significantly lower production levels and fewer returns, resulting in a substantial improvement in costs. If partial pooling by a sales agent produces results similar to the current system, it highlights that this approach alone delivers little variance reduction, making mid-week replenishment or a move toward fuller pooling essential to capture meaningful benefits. To gain buy-in across the organization, Assaf must communicate the operational and financial benefits of these changes clearly and with data. Demonstrating how reduced production and lower returns can free up printing capacity and improve overall company costs will be critical (i.e., by conveying the calculations and strategies demonstrated above). 7