SAP Analytics Cloud

Practical Analytics Chapter 12 | Exercise 3 | Edition 3

12.3

Global Bike Case Study

Nitin Kalé, University of Southern California

Nancy Jones, San Diego State University

IMPORTANT NOTE:

This case study requires data visualization, Cluster Analysis, and Forecast Analysis on SAP Analytics Cloud (SAC).
No software/academic assistance will be provided for this project. For data visualization, please review Chapter 6 slides and video tutorials in help documents of SAP Analytics Cloud (


Here is the Link

). For Cluster analysis on SAC, I strongly recommend reviewing/completing the optional case study “9.1. Segmenting Stores Using Clustering” in Module E of Canvas. For forecasting, you may want to review what you learned In Chapter 10 and HW 10.

Metadata

For your analysis, the data set (i.e., GB_AnalyticsData3.xlsx) includes variables related to Global Bike’s most recent sales data (see SalesdataAct tab in the dataset). You are supposed to prepare and load your datasets into SAP Analytics Cloud without any assistance.

The data set for your analysis includes variables related to Global Bike’s most recent sales data. You should be familiar with most of the variables, but here are some hints on this data set.

Problem Statement

The sales team at Global Bike (GB) is concerned about new competitors in the market. A strategic planning meeting is scheduled for the first of the month. Your job is to explore the data and provide insights on GB’s current customer base and sales data.

You need to complete the following analyses for the sales team. Be sure to include screenshots to support each of your answers and analyses. Submit your report as one Word document and name it as “Final Report_Your_Name.docx”. Bullet points or numbered items are acceptable for the report format. You will also share the SAC story as proof of your work. Your word document and URL(s) for your SAC story will be uploaded on Canvas. Bullet points or numbered items are acceptable for the report format.

Identify two (2) interesting and distinctly different relationships amongst three or more variables of the data set and discuss each briefly. You might hypothesize as to why the relationships are what they are. For example, “It makes sense that X is positively related to Y and Z because they are ….”

Create an appropriate visualization to illustrate the relationships you discovered. Use at least three (3) variables for each relationship (e.g., a Treemap showing Revenue and Gross Margin Ratio by Product) (e.g., a Marimekko Chart showing Revenue and Costs by Year for US Customers) (e.g., a Bar chart showing Revenue USD per Product Category and Country) (e.g., a TreeMap showing Revenue USD per Category by Country and City) (e.g., multiple line charts showing Count of Orders per CustDescr, Year) (e.g., a Bubble Chart showing Revenue per Order, revenue USD and other Per CustDescr).
Do not use the same three variables for both of your visualizations. Filter, sort and/or rank data if needed and explain what you filter if you do. You might even show the settings in your screenshot.
Be sure to identify the aggregation of your measures for each chart, perhaps in the title.

Explain to Nina why each of these visualizations is important to her team.

Example: (Treemap showing Revenue and Gross Margin Ratio by Product) This visualization shows the relationship of revenue and gross margin ratio by product. This is important so that Nina and her team can promote those items that not only bring in the most revenue but also those that have higher gross margins. It can also be used in decisions regarding what products to drop.

Nina Kane, her U.S. team, and her colleagues in Germany believe that Global Bike has 4 groups of customers with distinct characteristics. She also knows that customers have changed over the past few years and so any analysis that you provide on the groups should be focused on the newest data available. Create a cluster analysis for Nina

(Hint: I strongly recommend reviewing/completing the optional case study “9.1. Segmenting Stores Using Clustering” in Module E of Canvas)
. What can you tell Nina about the 4 customer groups? Use the following as a guide to help you answer this question.

a. Define how you clustered the transactions for Nina. What variables did you use? Did you use any filters? If so what were they?

b. Is there a group that stands out? Why? Take a screenshot of the cluster analysis and explain why you chose that cluster and any characteristics that define that cluster.

c. What can you tell Nina about the relationships of the members of each cluster and the clusters’ relationships to each other? Include screen shots of the visualizations you used to understand these relationships.

d. Explain other characteristics of the clusters by creating properly structured visualizations, (one visualization each), of the following.
Hint: be sure to use the cluster number as one of the dimensions in your visualization. Be sure to identify the aggregation of your measures for each chart, perhaps in the title.

i. Sales revenues by region and product category

ii. Profitability (gross margin ratio) by customer and division

The sales team is also interested in sales trends overall. Create a forecast of sales for the next 12 months.

Is the forecast reliable? Why or why not?

Save and share your story/results on your personal folder of SAC

Hints

You will need to do some minor data clean up and you will need to be sure that you have at least one date dimension.

Avoid using column or bar charts for every answer.

Grading Parameters

Did you choose the right model(s) for the question asked?

· Did you get the correct answers to specific questions?

Were you able to accurately explain the results of your model(s)?

Do your explanations show an understanding of the data in the data set?

Did you choose the best visualizations for the data represented?

· Did you choose the correct variables aggregated at the appropriate level for the question at hand?

· Are your visualizations properly formatted, (i.e.; titles and legends), and easily interpreted?

Nitin Kalé & Nancy Jones © 2024 2 of 3