Power Tool Segmentation Case

(1) Calculating willingness-to-pay:

In the 2-segment case, calculate the first consumer segment’s willing to pay for a top slider over a side slider switch. Report this willingness-to-pay. Show how you set up your calculations and discuss your findings.

• Step 0: we can see there are two sets of part worth utilities related values. One is Part Worth Utilities, the other one is Part Worth Utilities Rescaled for Comparability. I chose to use the latter one, because it’s easier to compare and see the difference.

The data set is as follow:

[pic 1]

• Step 1: Calculate the different utility value for a top slider over a side slider switch in 1st segment customers

99.01363 - 67.80978 = 31.20385

• Step 2: Calculate Dollar value of a util using the extremes of the part-worths for the price attribute

(38.32741 – (-20.51860)) utils = $129 - $79

                         1 utils = $50 / 17.80881

                                 ≈ $0.85

• Step 3: Calculate WTP for a top slider over a side slider switch in 1st segment customers

                = 31.20385 * $0.85

                ≈ $26.52

• Step 4: Findings. In two segment group statistical scenario, the 1st segment customers are willing to pay $26.52 for a top slider switch over a side slider switch angle grinder.

(2) Calculating attribute importance weights:

In the 2-segment case, calculate the first consumer segment’s importance weights associated with each attribute. Report the importance weights. Show how you set up your calculations, discuss your findings.

• Step 1: Calculate Attribute Utility Range (using Part Worth Utilities Rescaled for Comparability)

Utility Range = Highest Utility Value of an attribute - Lowest Utility Value of an attribute

For Brand Range is calculated as 58.22688-(-71.23369) = 129.46

For Price Range is calculated as 38.32741-(-20.51860) = 58.85

For Amps Rage is calculated as 48.20749-(-36.34262) = 84.55

For Life of Product Rage is calculated as 56.02714-( -76.08546) = 132.11

For Switch Rage is calculated as 99.01363-(-91.41781) = 190.43

For Grith Rage is calculated as 2.29964-(-2.29964) = 4.6

• Step 2 : Calculate Total Attribute Utility Range

Total Utility Range = ∑ Utility Range 
Total Attribute Utility Range = 600

• Step 3: Calculating Relative Importance of attributes

Relative Importance of attribute = (Attribute Utility Range/Total Attribute Utility Range)*100

[pic 2]

• Step 4: Discussing findings.

For the 1st segment consumer group in 2-group scenarios, they care about Switch very much, but not very price sensitive, they may sacrifice Price if the Switch is really what they want. And this segment customers ignore the Girth specification, so company can focus more on other attributes.

(3) Choose the optimal number of segments:

Reflect on the differences between the results across the different numbers of segments (i.e., 2, 3, 4, and 5 segments). What are some benefits to having higher numbers of segments? What are some of the main disadvantages of having additional segments? In both cases, be as concrete as possible in your discussions, illustrating advantages or disadvantages through specific examples. What number of segments best balances the pros and cons related to model fit and model usefulness? Briefly justify your decision.

First of all, one of the most essential consideration is how managerially relevant all the segments are. From a management perspective, if it makes sense to simplify or reduce the number of segments, to make it more usable, then I will use small numbers of segments, vice versa.

• Pros and Cons of a higher numbers of segments:

In this B&D case, I would like to look at the Model Fit Index numbers first, looking for smallest number, that is the one with the best fit between model fit and model complexity for each model.

[pic 3]

We can see: CAIC and BIC show that, from the statistic perspective, there are 3 segments in the market. And AIC shows 5 segments and ABIC shows 4 segments are in the market, all from statistical perspective.  And combining with managerial judgement, I think it makes sense to go with 3 groups for this case.

...