Segmentation Exercise:

This exercise builds directly off the Product Optimization Exercise we just did. The exercise calls for follow up work to characterize the segments. For this exercise, apart from the files from the Product Optimization Exercise, we need demographics data, which are provided in this file: demographics$-$full.xlsx$.$ The demographics variable are: income $($in thousands of

dollars per year$),$ age $($in years$),$ sports $(=1$ if the person is sports active, $0$ if not$),$ gradschl $(=1,$ if the person has a master's degree or beyond, $0$ if not$).$

If you did the Product Optimization Exercise correctly you have a file similar to this one: mugs$-$analysis$-$full$-$incl$-$demographics.xlsx

This file contains a correct execution of the core calculations in the Product Optimization

Exercise and also contains the demographics data of the demographics$-$full.xlsx file.

This exercise has two parts, the first on product affinity based segmentation and the second on classical segmentation.

Part $($A$)$ Product affinity based segmentation:

For this part, we need to assume a market scenario, which we will take to be the one below:

Incumbents

$1$: $$30,3$ hrs$,20$ oz$,$ Clean Easy, Leak Resistant, Brand A

$2$: $$10,1$ hrs$,20$ oz$,$ Clean Fair, Spill Resistant, Brand B

Our Candidate

$3$: $ $30,3$ hrs$,20$ oz$,$ Clean Easy, Leak Resistant, Brand C

Brand C$\text{'}$s product has the identical attribute level combination as Brand A$\text{'}$s product. You may think that Brand C$\text{'}$s marketing team is committing an error, but perhaps not because that attribute combination level is actually the one that maximizes expected profit per customer, as you may have discovered in the previous HW$.$ To build your intuition, you should reflect

on why it may be optimal for Brand C to mimic Brand A$.$

As in the previous exercise, we have the following attributes and attribute levels:

Price: $$30,$ $$10,$ $$5$

Time Insulated: $0\mathrm{.}5$ hrs$,1$ hrs$,3$ hrs

Capacity: $12$ oz$,20$ oz$,32$ oz

Cleanability: Difficult $(7$ min$),$ Fair $(5$ min$),$ Easy $(2$ min$)$

Containment: Slosh resistant, Spill resistant, Leak resistant

Brand: A$,$ B$,$ C

and the following cost structure:

Time Insulated: $0\mathrm{.}5$ hrs costs $$0\mathrm{.}5,1$ hrs costs $$1,3$ hrs costs $$3$

Capacity: $12$ oz costs $$1\mathrm{.}00,20$ oz costs $$2\mathrm{.}6,32$ oz costs $$2\mathrm{.}8$

Cleanability: Difficult $(7$ min$)$ costs $$1,$ Fair $(5$ min$)$ costs $$2\mathrm{.}2,$ Easy $(2$ min$)$ costs $$3\mathrm{.}0$

Containment: Slosh resistant costs $$0\mathrm{.}5,$ Spill resistant costs $$0\mathrm{.}8,$ Leak resistant costs $$1$

You are given data on the preference parameters of $311$ consumers in this file: mugs$-$preference$-$parameters$-$full.xlsx$.$ This is the same input file used in the Product Optimization

Exercise.

Your tasks for Part $($A$)$:

$1.$ Compute and report the characteristics of the affinity$-$based segment for Product $3($our candidate as given above$).$ Report the characteristics in terms of the following descriptors: IPr, Iin, ICp, ICl, ICn, IBr, I$*$pPr$30,$ I$*$pPr$10,$ I$*$pPr$05,$ I$*$pIn$0\mathrm{.}5,$ I$*$pIn$1,$ I$*$pIn$3,$ I$*$pCp$12,$ I$*$pCp$20,$ I$*$pCp$32,$ I$*$pClD$,$ I$*$pClF$,$ I$*$pClE$,$ I$*$pCnSl$,$ I$*$pCnSp$,$ I$*$pCnLk$,$ I$*$pBrA$,$ I$*$pBrB$,$ I$*$pBrC$,$ and the demographics income, age, sports and gradschl. For this you need to compute the weighted average of all the columns in the "for$-$cluster$-$analysis" worksheet, weighted by the probabilities given in column BF of

the "mugs$-$full" worksheet. You should not do this in Excel directly. You are to load the data into python or R and compute the weighted averages in R or python; this is to help you build familiarity with these languages and prepare you better for your job.

$.$ Repeat the step above for the product of brand A and the product for brand B$.$ Compute also the overall mean for each descriptor $($this done by a simple average of each descriptor across all $311$ customers$).$ Compute the log$-$lifts for all variables for the affinity based segment for each product and focus on the large positive and negative numbers to figure out how each segment is different from the other segments and the overall population average. $($You will use these in the next step to come up

with a verbal description that characterizes each segment.$)$

$3.$ Use your findings from Step $2$ above to come up with a verbal description and a persona story that characterizes each segment. The persona story gives a mental image to the marketing manager not only in terms of the descriptors in the dataset, but also in terms of plausible hypothesized characteristics that go beyond the descriptors available.

Steps $1$ and $2$ above are similar to what we saw in the lecture session's spreadsheet: mugsanalysis$-$limited$-$affinity$-$segmentation.xlsx$.$ The same spreadsheet, with cells highlighted for the analysis of the competitive advantage of product $3$ over product $2$ is here: mugs$-$analysislimited$-$affinity$-$segmentation$-$compare$-2-$for$-$customers$-$of$-3.$xlsx$.$ The PDF file summarizing the key results in a compact manner for the lecture's data $($not the HW data$)$ is given here: mugs$-$analysis$-$limited$-$affinity$-$segmentation$-$display.pdf$.$