Optimizing marketing campaigns
Statistical analysis with Excel of customers' attitude and characteristics in previous campaigns
This is an open project for the Maven Analytics Marketing Challenge.
I defined the business task after preliminary exploration of the dataset.
The business task
A hypothetical company specializing in food products recognizes the importance of precision in their marketing efforts. As they prepare for their next marketing campaign, they have gathered data from 2,240 customers, containing details such as customer demographics, product sales, campaign successes/failures, and channel performance. They want to use it to craft marketing campaigns that truly resonate, engage, and convert. The following business objectives are derived from this task:
​
What is the demographic profile of customers who accepted the offer in the last campaign compared to those who didn't?
​
Is there any product that can be used to predict the probability of accepting an offer through a marketing campaign?
​
What are the purchase behaviours associated with a positive response in previous marketing campaigns?
Key insights
The demographic traits of the most succesful target customers of the last marketing campaign were one or more of the following:
-
A person in their 30-40s (and to a lesser extent, in their 70s)
-
With higher education
-
With an average household income of 70-90K $ / year
-
Not in a romantic relationship
-
Living in Spain, Saudi Arabia, or Australia
Accepted offer in last campaign
​
Rejected offer in last campaign
3 demographic profiles showed a success rate >60%
PROFILE 1
-
30-50 years old
-
$80-100K / year
-
Married
-
Living in Saudi Arabia
PROFILE 2
-
PhD
-
$70-100K / year
-
Not in a relationship
-
Living in Spain
PROFILE 3
-
Graduate or higher
-
$80-100K / year
-
Not in a relationship
-
Living in Australia
Customers accepting the offer in marketing campaigns have spent on average more money on wine, meat, and gold over the last 2 years than those who didn't. These observations are supported by inferential statistics (heteroscedastic independent-samples t-test, p < .05)
p = 0.004
p = 0.018
p = 0.002
People that have been customers for 4 years or more have the highest percentage of offer acceptance, especially notable on the last (6th) marketing campaign. The majority of them have used only 1-3 discounts in purchases during that time.
DATASET LIMITATIONS
Unknown date for data collection. It is assumed it was collected on 31/12/2016
(see data cleaning log in downloadable Excel file for full explanation).
Data might be outdated (collection happened between 2012-2016)
24 blanks in income data → impossible to find this information,
so they were removed from demographic analysis.
Data-driven recommendations
By targeting specific demographics, aligning promotions with product preferences, and catering to long-term customers
Demographic customer segments
Segment marketing strategies to target individuals aged 30-40s and those in their 70s, especially in countries like Spain, Saudi Arabia, and Australia.
​
Tailor campaigns to resonate with individuals with higher education, average household income of $70-90K per year, and those not currently in a romantic relationship.
​
Concentrate sales team efforts on demographic profiles with a success rate above 60%.
Product preferences
Craft marketing campaigns that emphasize wine, meat, and gold products.
​
Design targeted promotions for these product categories.
​
Concentrate sales team efforts on customers with high expenditure on any of these product categories.
Loyalty and discount use
Capitalize on higher acceptance rates by providing tailored incentives to long-term loyal customers. For example, design loyalty programs that provide exclusive benefits to customers using discounts.
Concentrate sales team efforts on customers who have been with the company for 4 years or more.