Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
MANAGEMENT
PGP TERM IV
CASE STUDY - WILLS LIFESTYLE
For all analysis assume that the margin associated with internal production is 30% (of cost) per style and
the average cost of overstocking is 10% (of cost). Thus, a shirt that retails at Rs. 1300 has a cost of Rs.
1000 and a cost of overstocking of Rs. 100. Use the Demand data in Exhibit 7.
Q1a. Consider the case when LRBD sources from a third party (as was the case prior to 2003). The third
party has a unit cost which is Rs. 50 per style lower than the cost implied in Exhibit 7. The retailer has
to place a single order (before the selling season, obviously) and a minimum lot of size of 2000 units per
style. What would be the order size for each of the style? Assume Wills Lifestyle will place orders for all
style (it cannot be zero for some style).
This problem is a direct application of newsvendor model where the optimum order quantity for each
style is calculated based on cost of understocking and cost of overstocking and demand. Since, the order
quantity must be of minimum lot size of 2000, the final order quantity is given as max (2000, Optimal
order quantity).
Cost of understocking: Sales Price – Cost Price = Sales Price – Third Party Cost Price
Q1b. What happens if the supplier only offers a total capacity of 25,000 units but there is no minimum
lot size per style?
This is an optimization problem, with the objective function being maximization of expected profit. The
formulation for the same is given below as:
𝑭𝒊𝒏𝒅 𝑞1 , 𝑞2 , 𝑞3 , 𝑞4 , 𝑞5 , 𝑞6 , 𝑞7 , 𝑞8 , 𝑞9 , 𝑞10
𝑞1 , 𝑞2 , 𝑞3 , 𝑞4 , 𝑞5 , 𝑞6 , 𝑞7 , 𝑞8 , 𝑞9 , 𝑞10 𝑎𝑟𝑒 𝑖𝑛𝑡𝑒𝑔𝑒𝑟𝑠
Where 𝑞1 is the quantity to be ordered for Sun Orange and so on. 𝑞1 , 𝑞2 … . . 𝑞10 are the decision variables.
The objective function is to maximize expected profits, which can be written as:
𝐸(𝑠𝑎𝑙𝑒𝑠) = 𝑢1 − 𝜎 ∗ 𝐿(𝑧1 )
1|Page
𝑴𝒂𝒙𝒊𝒎𝒊𝒔𝒆: 𝑟1 𝐸(𝑠𝑎𝑙𝑒𝑠1 ) − 𝑚1 ∗ 𝑞1 + 𝑟2 𝐸(𝑠𝑎𝑙𝑒𝑠2 ) − 𝑚2 ∗ 𝑞2 + 𝑟3 𝐸(𝑠𝑎𝑙𝑒𝑠3 ) − 𝑚3 + 𝑟4 𝐸(𝑠𝑎𝑙𝑒𝑠4 )
− 𝑚4 ∗ 𝑞4 + 𝑟5 𝐸(𝑠𝑎𝑙𝑒𝑠5 ) − 𝑚5 ∗ 𝑞5 + 𝑟6 𝐸(𝑠𝑎𝑙𝑒𝑠6 ) − 𝑚6 ∗ 𝑞6 + 𝑟7 𝐸(𝑠𝑎𝑙𝑒𝑠7 ) − 𝑚7
∗ 𝑞7 + 𝑟8 𝐸(𝑠𝑎𝑙𝑒𝑠8 ) − 𝑚8 ∗ 𝑞8 + 𝑟9 𝐸(𝑠𝑎𝑙𝑒𝑠9 ) − 𝑚9 ∗ 𝑞9 + 𝑟10 𝐸(𝑠𝑎𝑙𝑒𝑠10 ) − 𝑚10 ∗ 𝑞10
Where mi is manufacturing cost of ith product, ri is revenue of ith product
Given the capacity constraint of 25,000 units, the summation of all the quantities has to be less than or
equal to 25000.
When the capacity constraint was set to 25,000 units, it is interesting to note that the profit declined to
3,010,680 . This means it is better to let some capacity go underutilized for better profit.
Q2. Now consider the process post-2003. LRBD divides the sales period or season (for example, summer
selling season) into six sub-periods and arranges for multiple replenishments after bringing in an initial
order quantity. Assume that each sub-period averages about one-sixth of the season’s demand. What
initial order quantity (for each style) should LRBD place (and in successive order periods)? The minimum
lot size is 800 units per style.
On dividing the sales season into six uniform periods, the problem essentially becomes a newsvendor
model for each period. But since they follow demand replenishment model so initial quantity will be the
optimum quantity given the minimum lot size. The final quantity to be ordered for each subsequent period
would be optimal order quantity minus the left-over inventory of last period to replenish the stock. The
conditions for each period would be that if order quantity is less than last period’s inventory, the final
2|Page
quantity order would be completely from left over inventory as well as the freedom to choose to not order
any quantity.
Also, since the sales divided into six periods, the means sales would become total mean sales/6 while the
standard deviation for each period would be Standard Deviation/ √6 .
3|Page
4|Page
2. Assuming no minimum lot in periods following the first purchase.
The quantity to be ordered is given below as:
5|Page
6|Page