Estimate the Lift from Each Price Cut

1.

a. From the historical data, estimate the lift from each price cut (i.e., the average increase in weekly from each price reduction). Also, estimate the average demand at each price point. (Caution: remember that demand is not the same as observed sales.)

[pic 1]

The lift from each price cut has been calculated by the average increase in weekly from each price reduction by not including any periods with sellouts in estimating the demand at the discount price.

To estimate the average demand at each price point, we applied regression analysis on the price reduction and increase in sales. We found the following relationship;

[pic 2]

Using this results, we can calculate the estimated demand at each point (e.g. estimated demand for price $54 is 102.03 = 406.98 + (-5.666 x 54)).  

b. Estimate the variability in sales from one item to another (i.e., the standard deviation on the average weekly sales at a given price point across items). Also, estimate the variability in weekly sales for a given item (i.e., the standard deviation on weekly sales within a single item).

[pic 3]

c. If demand was not variable and followed the average lift for each price point, using your estimates in (a), find an optimal pricing strategy for the best price to charge in each week of the sales season.

Please refer to the table below. Using the calculated lift, we estimated the demand for each price. Then, by using deterministic markdown linear model, we figured out an optimal pricing strategy.

 [pic 4]

[pic 5]

D. With uncertain demand, how should your strategy change compared to the strategy from (c)? Describe what the strategy should attempt to do and how it should use the demand information in general.

The most important thing in the pricing strategy when the demand is not variable would be determine when and how much you lower the price. As there is no cost associated with a loss of sales when demand cannot be met, the pricing strategy would be much easier.