Optimized Analytics – Final Project
Report (50% of grade)
The report must answer all parts of the prompt below and should be submitted as a .zip file which contains:
• Well-commented and neatly presented R or Python code (or combination of R and Python).
• Word document (1-2 pages) providing the answers to each of the questions in the prompt, along with any necessary explanation. CSV file providing the answer to #3 in the prompt
Marketing Optimization
Your company is looking at optimizing their social media marketing by investing in various
advertising campaigns. There are 250 Instagram campaigns and 250 Facebook campaigns
available. Each runs for one month and reaches 30,000 people. Revenue for each campaign is found as numberOfCustomersReached*conversionRate*customerPurchaseAmount. Profit for each campaign is found as Revenue – Cost.
1: Decide how much money to spend on the different ad campaigns to maximize total profit
while meeting a budget of $2,000. Assume you can invest a partial amount (however small
you would like) in each campaign, but you can’t run each campaign more than once. Return
the number of campaigns run, total cost, total profit, average percentage reach to existing
customers, average percentage reach to new customers, and average percentage reach to
female customers.
2: Decide how much money to spend on the different ad campaigns to maximize total profit
while meeting a budget of $2,000. Each campaign is all or nothing – you can either run the
whole campaign for the full cost or not run it at all. Return the number of campaigns run, total
cost, total profit, average percentage reach to existing customers, average percentage reach
to new customers, and average percentage reach to female customers.
3: You’re not sure how much your budget requirement is hurting your marketing efforts, so you would like to see how sensitive your results are to your budget. How profitable will you be if you have a higher or lower budget? You would like to evaluate every budget option in $500 intervals between $500 and $10,000 (i.e. $500,$1000,$1500, etc.). Assume you can invest a partial amount (however small you would like) in each campaign, but you can’t run each campaign more than once.
#
BudgetThreshold
NumberOfCampaigns
ExpectedReturns
1
500
4.913978
17.88408
2
1000
9.185185
77.8219
3
1500
13.35294
157.5278
4
2000
18.73611
258.8989
5
2500
23.04132
374.9837
6
3000
27.82143
524.1468
7
3500
32.00775
705.085
8
4000
36.37748
895.4824
9
4500
40.86076
1113.708
10
5000
45.4545
1362.384
11
5500
50.74713
1655.17
12
6000
57.73188
1982.112
13
6500
63.52258
2364.021
14
7000
68.38095
2785.029
15
7500
73.45614
3236.58
16
8000
78.99286
3710.537
17
8500
84.38849
4210.68
18
9000
89.544
4719.351
19
9500
94.15385
5257.95
20
10000
99.69531
5819.14
4: To evaluate the different budget options, write a function which takes in budgetThreshold as an input and outputs the number of campaigns run, total cost, total profit, average percentage reach to existing customers, average percentage reach to new customers, and average percentage reach to female customers. Return these values in a csv file.
The csv file should have 20 rows and the columns should be:
a. budgetThreshold
b. numberOfCampaigns
c. totalCost
d. totalProfit
e. mean%reachToExistingCustomers
f.mean%reachToNewCustomers
g. mean%reachToFemaleCustomers