Business Analytics PROJECT 1: DESCRIPTIVE ANALYTICS
SPRING 2019
The following two cases reflect some past work completed by data analytics team at Element451. As such they are used for educational purposes, to teach the challenges of data curation, integration, and insight generations. Data has been anonymized but participants cannot share this datasource outside the scope of this class.
Use Case 1: Email Engagement Analysis
Data: https://drive.google.com/drive/folders/1fnzk10qSU3ttFBpCT_QV5IQKDA6sJe_1?usp=sharing
Your university is trying to answer questions on email engagement. The school is worried about the perception of email spam, and want to reduce the number of emails sent to students.
To focus on email engagement means that you are looking at how students interact with email from a school. For example, do they:
● Open the email?
● Read the content?
● Click on links?
● Follow a call to action?
For this project, your team needs to answer the following questions:
● What is the average number of emails that we are sending per student?
● What is the average open rate per student?
● How many students (count and %) are not opening emails?
● How many emails does it take (number of delivered) to get students to open an
email.
● Among students who unsubscribed (emailUnsubscribed), how many emails on
average were they sent?
● How does email engagement change over the course of enrollment milestones
(Lead→ Prospects→ Applicants→ Admitted)
Use Case 2: Persona Modeling
Data: https://drive.google.com/drive/folders/1Vt3DRH_jLYWVoGUQZsc5z-tEhH0zkj58?usp=sharing
Targeting prospective customers continues to be a resource-intensive effort in marketing operations across the country. It usually requires large budgets, endless personnel time, and multiple vendors to identify and message the right client. These days, marketing and advisement professionals are turning to the technique of Marketing Personas as a way to run effective and successful marketing strategies.
Sample of an marketing persona
Persona development, as a design tool, was first introduced by Alan Cooper in 1998 as a
way to “a precise description of our user and what he wishes to accomplish”. According to Cooper a persona is “a semi-fictional representation of your ideal customer based on market research and real data about your existing customers. Ultimately, in persona development marketers are aiming to represents the goals, motivations, and behaviors of a target user base. So, in a nutshell, a marketing persona is a composite sketch of a key segment of an audience in which we encode/decode customer experience so that:
● We can better understand clients needs/interests
● Develop knowledge of where students motivations and aspirations
● Merge targeting & messaging
For this project, your team is required to produce personas for the Marketing Department of a US based University. The University is looking to optimize a marketing budget of $800k to generate a healthy and qualified pool of applicants. As such, your persona modeling should include the following:
● Modeling using only summary statistics
● A minimum on 5 persona profiles
● Normalization of data to account for outlines and unbalanced variables
For this project, your team needs to answer the following questions:
● What is causing a student to engage with the University?
● Which sources are more effective?
● How many sources does it take to move a student through the enrollment funnel?
● What is the best way to market/recruit these students?
Submission, Deliverables, & Grading
Project is due in class and defended on March 5. Each team will submit slides in NYUClasses and provide a 7-10 minute presentation followed by a 5 minute Q&A. Team should seek to answer the following questions when presenting:
● Overview of the problem – what is the problem you are trying to solve. How is this analysis relevant to your client
● Overview of data: what was the final dataset used for the analysis?
● Analysis & Results: review of techniques used and results from analysis. What
patterns emerged from the data? Are there any anomalies in your use case? What
associations emerged?
● Dashboard: visualize your primary findings in an executive dashboard. The
dashboard is the “take away” from the presentation. Basically, the client will take the dashboard to review the analysis you presented. (HINT: The analysis above includes primary and secondary findings. The dashboard is an executive report, thus it should only have a subset of your primary findings)
Grading will be based on the following:
● Quality of presentation – 20%
● Quality of analysis – 50%
● Quality of dashboard – 30%