Assessment scenario: Students in BUSML 4202 analyze a large sample set of customer satisfaction data using a multi-stage process that enables them to experience three distinct approaches: manual coding, computational text analysis, and prompting large language models to process the data. Learn how using AI as a collaborator (rather than a shortcut) expands students' understanding of AI’s role in the research process while preparing them for their future careers.
Instructor: Jeffrey P. Dotson
Department: Marketing and Logistics, Fisher College of Business
Course: BUSML 4202: Marketing Research
Contact: dotson.83@osu.edu
Course Details
Course format: In-person
Typical enrollment: 50
Instructional staffing: One primary instructor
Learning context: Junior or Senior Students in the Marketing specialization in the Fisher College
Expected Learning Outcomes
- Evaluate strengths and limitations of manual, statistical, and AI-based text analysis
- Assess AI system inputs (data quality, prompt construction) and outputs (coherence, accuracy, potential bias)
- Engage critically with generative AI models through iterative refinement rather than passive acceptance
- Translate analytical findings into strategic recommendations appropriate for professional consulting contexts
Opportunity
Engaging AI as a collaborator, not a shortcut
BUSML 4202 introduces students to data-driven decision-making in marketing, covering the full research process from qualitative inquiry and survey design to data analysis and strategic recommendations. As generative AI has transformed professional marketing research practice, the course has evolved to help students understand how tools like ChatGPT, Claude, and Gemini can augment every phase of the research process—not as shortcuts, but as collaborators that enable analysts to focus on strategic interpretation and decision-making.
Change
Understanding AI's growing role in the research process
Students are asked to assume the role of marketing research consultants hired by a private equity firm evaluating a potential acquisition. Students are provided with an instructor-created description of the target company and sample data from a customer satisfaction survey. The data includes responses to likert-scale questions and over 130 pages of open-ended feedback.
Students are instructed to complete a multi-stage process for analyzing the sample data, which is designed to help them understand AI’s role in modern research.
In stage 1, students manually code the data. They begin by reading a subset of comments, experiencing firsthand the time demands, subjectivity, and interpretive challenges of traditional qualitative analysis.
In stage 2, students apply computational text analysis techniques (topic modeling or word clouds), learning the complexities of data cleaning, model fitting, and thematic interpretation inherent in bag-of-words approaches.
In stage 3, students use large language models to process the full data set. Through prompt engineering, they define the model's role (e.g. "act as a marketing research consultant"), specify objectives (e.g. "identify major customer themes"), and request evidence-based outputs (e.g. "select representative quotes"). They iterate on prompts to refine themes and develop actionable solutions.
After completing these three stages of analysis, students are instructed to: identify the top three customer issues emerging from the unstructured text data; provide verbatim customer quotations illustrating each issue; propose actionable solutions to address identified problems; and synthesize their findings into a two-page executive report for client delivery. I instruct students that they can use LLMs for any and every state of this analysis and writing process, and that–at the end of the day–they are responsible for the work that they turn in.
Reflection
Preparing students for the profession
This assignment prepares students for a profession where AI tools are increasingly standard while emphasizing that technology amplifies rather than replaces human expertise. By experiencing the full analytical spectrum—manual, computational, and AI-augmented—students develop informed perspectives on when and how to deploy AI effectively in marketing research contexts.
The exercise consistently reshapes student perspectives: Students begin by seeing AI as a shortcut—but they leave recognizing it as a collaborator. The goal isn't to let the machine think for you, but to focus your attention where human judgment matters most: interpreting results, identifying implications, and designing better decisions.
This reframing—from labor-intensive processing to strategic interpretation—mirrors the transformation occurring in professional marketing research. Students emerge with both technical competency in AI-assisted analysis and a nuanced understanding of how to manage and leverage these systems responsibly.