Using GPT-3 and Hacker News for slightly creepy market research
12 min read
"What if I fed GPT-3 a community's online conversations? Could I use that to simulate interviewing a community member for market research?" That was the spark.
I know, I know, developer writes an article about avoiding interacting with people...but that's not what this is. I start with a poorly thought-out goal and end with ok results (You can find the results just below). Still, along the way, I learn a bit about the general problem of computational market research. Here's a tl;dr:
- Analysis that you don't trust or understand will not affect your uncertainty. Your personal context affects what techniques are most effective.
- The two pillars of computational market research are: High precision search functionality and SQL as a query interface. They help solve the most fundamental analysis problem: What data contains the answers to my questions?
- Your interpersonal skills and intuition have high transferability to marketing.
Act 1 : How do I wield GPT-3?
Act 2 : How do I do useful research?
Act 3 : How do I answer a question about the market?
-  Hewson, C. and Laurent, D. (2012). Research Design and Tools for Internet Research.In: Hughes, J ed. Sage Internet Research Methods. Sage.
-  Miles, M. B., Huberman, A. M., & Saldaña Johnny. (2013). In Qualitative Data Analysis: A methods sourcebook (p. 36). essay, Sage.
-  Grimmer, J., Roberts, M. E., & Stewart, B. M. (2022). Principles of Discovery. In Text as data: A new framework for Machine Learning and the Social Sciences. essay, Princeton University Press.
-  Zhao, S., Grasmuck, S., and Martin, J. (2008). "Identity construction on Facebook: Digital empowerment in anchored relationships." Computers in Human Behavior, 24(5), 1816-1836.
-  Marwick, A. and Boyd, D. (2011). "To see and be seen: Celebrity practice on Twitter." Convergence, 17(2), 139-158.
-  Ignatow, G., & Mihalcea, R. (2018). In An introduction to text mining: Research Design, data collection, and analysis (p. 58). essay, SAGE Publications, Inc.
-  Kozlenkova, I.V., Samaha, S.A., and Palmatier, R.W. (2014) ‘Resource-based theory in marketing,’ Journal of the Academy of Marketing Science, 42(1), pp. 1–21.