Profit Analysis

Midterm ProjectYou have recently begun an internship for an online office supply store, Office Mate. They have
been in business since 2012 and have done well. However, they are beginning to face more
competition from Amazon and several other online retailers, so they need to determine how to
become more profitable. Your manager, Quian Xu, has come to you and a co-worker to complete
an analysis and come up with two recommendations to improve profitability. The specifics of the
task are outlined below. Quian will be sharing your findings with upper level management and
the board of directors, so it is critical that your writing and graphs are clear to people without a
technical background. Analyze the data, determine two significant findings (practical and
statistical to be showed in examples) related to profitability, create a professional quality
visual(s) for each finding, and provide a brief write up of your findings and recommendations in
a memo using best practices. Use the data “mtp_off_mate.csv”, which includes the following
variables:
1. Order ID: Each order ID is unique, but may contain several products
2. Order Date: Date the order was placed
3. Ship Date: Date order was shipped
4. Ship Mode: Shipping method
5. Customer ID: Each customer has a unique ID, but can make more than one order
6. City, State, Postal Code: For the location of each sale in continental US
7. Region: Sales region in the United States (Central, East, South, West)
8. Product ID: Product ID code
9. Product Name: Name of product sold
10. Segment: Customer segment (Consumer, Corporate, Home Office)
11. Category: Type of product sold (Furniture, Office Supply, Technology)
12. Sub-Category: Detailed type of product sold
13. Revenue = Price * Quantity (though the dataset does not include Price)
14. Quantity = number of units of the specific product sold for that order
15. Discount: Percent discount from full Price. Large discounts are used to clear inventory.
16. Profit = Revenue – Cost (not including shipping or tax, both passed directly to customer)
Each observation is for the sale of a specific product from an order that may contain multiple
products. The Order ID variable links products from the same order. The customer ID identifies
unique customers that may have multiple orders
Word Memo Format: 2-page maximum including graphs and/or tables with appropriate
memo header for a non-technical audience. Your memorandum should include the
following: (Please see the Word memo code and result example, write the similar format
as it shows, use yliu@gmail.com for further contact info. )
1) An introductory paragraph giving background, brief description of the data, and brief
description of your findings. This paragraph is meant to engage your audience so it should
make the analysis compelling—why is the problem important to Quian and Office Mate.
2) A paragraph describing each of your two most important findings, each supported by a
visualization (and possibly a small table) that are explained in non-technical terms.
3) Conclusions and suggested course of action in the form of a recommendation, and possibly
other minor findings.
Technical Appendix: I would highly suggest to do the Technical Appendix as the two
example shows
4)The technical appendix contains your base, detailed, and statistical EDA, and the visuals to
load to your memo. It is important that your work is clearly organized (use a table of contents
(TOC)), easy to understand (code clearly documented), and reproducible (all in R, no cutting
and pasting). Title each step of your analysis so it appears in the TOC. The files must run
from the original data sets, so all data wrangling must be contained in the RMD files. The
technical appendix explains what you are doing and why, technical jargon is acceptable
because the audience is expected to have expertise. Write out what you observe and
questions that arise after each step during your base EDA – these are the data
comments/questions that you will examine in your detailed EDA. Think of each code chunk,
output, and your written observations as a page in your notebook. Test your findings
statistically (see examples), create your professional visuals, and save them externally.
Submissions: Submit the two html and two RMD files —title files using the following method:
“LastName_TA.RMD”, “LastName_TA.html,
“LastName_memo.RMD, and “LastName_memo.html.
Also include your name in the TA and memo headers.
Remember, all writing in the memo must be clear, concise and accurate – no repetition or
excessive use of adverbs and adjectives. No fluff or filler, every sentence and word should have
purpose. Quian does not have time to guess at what you have done, and she is not a data
analyst, so the memo must be written clearly in plain English with NO technical jargon.
However, the technical appendix can include jargon, but must be clearly documented so readers
do not have to guess what you have done and why.
Category
Introductory Paragraph – (3)
Score
Compelling (1), data (1), brief finding (1)
Finding 1 – (8)
Visual quality (4), descrip (2), measure (1), exceed (1)
Finding 2 – (8)
Visual quality (4), descrip (2), measure (1), exceed (1)
Conclusion – (3)
Summary (1), recommend (1), additional (1)
Technical Appendix – (24)
Clear organization (2), documentation of code (3),
data comments/questions (4), data wrangling (2),
base EDA (4), detailed EDA (3), stat EDA (3),
exceed expectations (3)
Mechanics of memo and TA – (4)
Grammatically (1) and typographically (1) correct,
effective memo design (1), no cut & paste (1)
Total
/50
Comments