BU MET AD654 Fall 2023

Marketing Analytics Fall 2023

©2020 Team Course Development

AD654: Marketing Analytics covers a wide range of descriptive, predictive, and prescriptive analytics techniques. Students in this course perform consumer segmentation, customer lifetime value analysis, A/B Testing, customer churn prediction, and other algorithmic processes using Python libraries such as scikit-learn, pandas, statsmodels, and scipy. In addition, AD654 also incorporates many types of data visualization tools, including both Tableau and the matplotlib and seaborn libraries in Python. Using the fictional backdrop of the amusement park Lobster Land, students build and assess data models that predict specific consumer behavior and help management to optimize decision-making processes regarding the 4 Ps of the marketing mix -- Product, Placement, Promotion, and Price. For the end-of-semester project, students will demonstrate several of the analytical skills taught in the course in order to develop recommendations for Lobster Land management regarding a series of decisions related to a business opportunity. [4 credits]

Prerequisites: AD571, edX-based pre-analytics laboratory ADR100, edX-based labs MwaM & SwaM

This course enables students to develop experience in the following areas:

1. Theoretical and practical understanding of core data mining concepts, techniques, and business applications

2. Systematic approach to framing and solving business analytics problems with the help of data mining methods and techniques;

3. Ability to identify the right data mining tools and techniques for various business analytics problems;

4. Hands-on experience in using the most popular business analytics and data mining tools and preparation for applying for job positions where familiarity with those tools is required.

Throughout our journey together in AD654, I will ask you to adhere to the following set of course norms:
Assume the best: Some things in this course will be simple; others will be difficult. Rest assured that nothing is intentionally designed to trick you. I’m not perfect and I can make mistakes in haste. If I post a file that says “Assignment #2 prompt” and it’s actually my grocery shopping list, then please assume that I have made a careless error, and not that it’s part of some elaborate scheme to trick you.
Monday Blackboard Announcements: Every Monday, I’ll make an announcement in Blackboard that will include a few bullet points that mention the topics that we’ll explore that week, along with upcoming due dates.
We’ll Always Have a Break in the Middle: Our class period is long. It will never consist solely of me lecturing in “one-to-many” format. We will always take a break of approximately 10 minutes, typically around the midway point of the period.
We’ll Always Start on Time. Slides will always be posted prior to start time: Class will begin promptly, at the officially-stated start time. Slides for each class will be posted prior to the start, so that you can follow along on your laptop if you wish to.
< 24 hours turnaround on e-mails, Homework turnaround by next due date: I will never go more than 24 hours without checking my BU e-mail. I will respond to all student e-mails in less than 24 hours. I will sometimes respond much faster than this, but please allow up to 24 hours. I will aim to have all submitted homeworks will be graded with comments prior to the subsequent assignment’s due date. I will sometimes grade homeworks in batches of 2 or 3, so if your friend’s homework was graded yesterday but yours still isn’t, there is no need to panic.
Effort matters: I realize that for many students, AD654 requires some steps outside of the “comfort zone.” Students who maintain a positive attitude and who put forth a strong effort tend to do very well in this course, regardless of what their knowledge level was on Day 1.
Missing a class – not a showstopper. Be an adult, get caught up: Life happens. Between job interviews, illnesses, family events, etc. you might miss a class. If you do, you can review the material by checking the slides on Blackboard. You can seek me out for extra help with assignments. “I missed that class” is not a valid excuse to simply not complete an assignment. 
If you see a msitake? Let me know. Remember rule #1.
If it ain’t raining? We ain’t training! Things can always go wrong, and sometimes when you least expect them to. My computer could decide on a forced Windows update 2 minutes prior to class starting. If/when any of these things occur? We won’t miss a beat -- we can always adjust, adapt, and overcome.

Page, Greg, and Huey Fern Tay: Marketing Analytics at LobsterLand: A Python-Based Approach. This book is available at lobsterland.net. It is free of charge.

Several videos related to data mining, machine learning, Tableau, and the Python language are available on the class Blackboard page. Most of the videos can be considered optional -- they are there as a learning aid that you can feel free to use at your discretion. Prior to quizzes, I will specify which videos are considered “inside the scope” of the quiz, if applicable.

For directions to get free remote access to our BU MET Virtual Labs, please visit:

http://www.bu.edu/metit/pc-labs/virtual-labs/

Anaconda Navigator (the program that we will use to launch Jupyter Notebook, our Python reporting tool) is installed in the Virtual Lab and available for use by any AD654 student. Tableau is available in the virtual lab as well.

Your performance in the course will be graded in the following areas:

Attendance, Participation, and Professionalism	10%
Quizzes (3 total)	50%
Individual Assignments (5 total)	20%
Group Project (Written Submission)	15%
Final Presentation	5%

Additional details for each grading component are provided below:

Quizzes: Each of the three quizzes will consist of 15 questions that students will complete during a 60-minute block of time. Quizzes will be open-note and open book. I will say more about the quiz format during the semester.

Assignments: Each assignment will evaluate students’ ability to apply selected marketing analytics techniques to the course material. The assignments will be graded based on a combination of accuracy of the analysis and quality of the report. More specific information for the format

and the contents of the assignments is available on the course Blackboard page.

The business running case concerns a fictitious theme park called Lobster Land. Set just outside of Portland, Maine offers many forms of entertainment to families from around the world. Through the backdrop of the Lobster Land case, students will analyze and solve descriptive, predictive, and prescriptive marketing questions. Such questions are as wide-ranging as “How can we better understand the demographic profile of big spenders at the Lobster Land arcade?” to “How can we incorporate k-means clustering analysis into a decision about expanding some aspect of the park?”

The team project is to apply the learnings in class to identify a data mining problem that can be solved with business analytics methods, models, and applications covered throughout the semester. Students will work in teams on this project. More information about this project will be made available on Blackboard midway through the semester.

Each team will deliver a 15-minute presentation during our last class session. More specific information for the format and the content related to this presentation can be found in the Project folder on the course Blackboard site.

All work requests from the instructor (quizzes, assignments, contributions in the teamwork, etc.) have due dates. These are the last dates that stated material is due. This means that it is a good idea to set personal targets before then as your personal completion date to avoid difficulties.Dates are often viewed by students as the date to turn in an assignment. We view assignment due dates as the last date on which to turn in an assignment. With this caution, please note that we are not inclined to accept late work; if late work should be accepted it will be done only after considerable weighing of rationale, and with penalty.

Students are expected to adhere to the highest standards of honesty and integrity for this course. University policy on academic integrity will be followed to the fullest. Students are encouraged to review the university policy on academic integrity including a detailed listing of activities warranting sanction. Anyone who fails to adhere to these requirements and/or otherwise engages in unethical behavior (including cheating on exams, false representation of self or one’s work efforts, use of unauthorized aids, etc.) will be referred to university administration for further action. In particular, the university's policy and consequences regarding plagiarism are clearly described in the official Boston University documents, and will be enforced without any compromises.

If you have a disability and will be requesting accommodations for this course, please inform the instructor early in the semester. Advance notice and appropriate documentation are required for accommodations.

Satisfaction of Department-Wide Goals

#	Goals	Category	Compliance
1	Critical and innovative thinking	Substantial	With the help of the assignments and individual exercises, students are expected to learn and choose the appropriate data mining model for problem solving and decision making.
2	International perspective	Some	The examples discussed in some data mining approaches and modules are applicable to both national and international organizations.
3	Communication skills	Substantial	Students are expected to participate in weekly group discussions, which support the development of communication skills.
4	Decision making	Substantial	Quantitative decision making is emphasized throughout the course.
5	Technical tools & techniques	Substantial	The course introduces a variety of tools and techniques including MS Excel based Frontline Analytic Solver Platform and R One.
6	Research skills & scholarship	Substantial	The course asks students to complete several assignments. In each assignment, students are asked to construct data mining models and apply decision support tools.
7	Professional ethics & standards	Substantial	The importance of professional ethics and standards emphasized throughout the weekly discussions.
8	Creative & effective leaders	Substantial	Understanding data mining and other business analytics models and using them for decision-making is critical for becoming creative and effective leaders

Class Date:	Lectures & Topics	Readings (from text)
06SEP	Topic 1: Course Overview, Intro to Marketing Analytics, Intro to Lobster Land; Python as an analytics tool	Ch. 0
13SEP	Topic 2: Market Insight & Market Sizing Analytics; Data Visualization with matplotlib and seaborn	Ch.1, 2
20SEP	Topic 3: Analytics for Market Segmentation	Ch. 3
27SEP	Topic 4: Product Analytics & Product Portfolio; Ratings-Based Conjoint Analysis	Ch. 4
04OCT	Topic 5: Customer Lifetime Value, Brand Metrics, and Digital Marketing Metrics	Ch. 5
11OCT	Topic 6: Experiment Design, A/B Testing, and Statistical Distributions	Ch. 6
18OCT	Topic 7: Understanding Classification Models and Assessing their Performance	Ch. 7
25OCT	Topic 8: Logistic Regression	Ch. 8
01NOV	Topic 9: Random Forests – Building an Ensemble Model with scikit-learn	Ch. 9
08NOV	Topic 10: Pricing	Ch. 10
15NOV	Topic 11: Forecasting	Ch. 11
22NOV	Thanksgiving Holiday – No Class
29NOV	Topic 12: Extracting Data from the Web, Text Analysis	Ch. 12, Ch. 13
06DEC	Topic 13: Recommender Systems & Advanced Modeling Techniques	Ch. 14, Ch.16
13DEC	Semester Presentations & Lessons Learned Team Project Write-Ups and Slides Submitted by 11:59 p.m. on 12DEC

Assignment #1: Due by 11:59 p.m. Sunday, 24SEP

Assignment #2: Due by 11:59 p.m. Sunday, 08OCT

Assignment #3: Due by 11:59 p.m. Sunday, 29OCT

Assignment #4: Due by 11:59 p.m., Sunday, 12NOV

Assignment #5: Due by 11:59 p.m., Sunday 03DEC

Quiz #1: Wednesday, 04OCT

Quiz #2: Wednesday, 01NOV

Quiz #3: Wednesday, 06DEC

All quizzes are open-note, but without any smartphone or computer usage. I will say more about this in class. Quizzes are a strictly individual effort.