INTEREST CATEGORY: MARKETING STRATEGY
POSTING TYPE: Journal News

Author: Min Ding

---

Call for Papers

Research and Replication Using Generative AI

Customer Needs and Solutions

When OpenAI launched its large language model, ChatGPT, on November 30, 2022, it received immense attention immediately.[1] While the true impact of Large Language Models (LLMs) and other Generative AI on academic research in business and marketing will be evident in the months and years to come, it seems apparent that it will fundamentally change the way we conduct academic research.

We believe that our discipline (journals) should embrace Generative AI and treat it as a new powerful tool to be added to our toolbox. To jumpstart Generative AI facilitated research, this call for papers by Customer Needs and Solutions invites scholars to replicate existing marketing work using Generative AI tools,

Goals

• Investigate what Generative AI can do in marketing, in terms of replacing human (academics or practitioners) for completing specific marketing tasks.
• Publish and synthesize the best practice of using Generative AI for completing specific marketing tasks.
• Explore potential limitations and biases of such practices.

Target Work (Work to be replicated):

• Academic:
  • Either a complete article (e.g., a review article) or a self-contained section of a published work (e.g., the literature review or introduction). Ideally, it should be a (section of) well cited article published in a top marketing journal.
• Practice:
  • A specific marketing practice that a firm has already used, such as a specific print advertisement, a video ad, an audio ad, a press release, etc.

Note the target to be replicated should not be in the training data of the specific generative AI tool used (e.g., Chat GPT, as of now, were trained with data before September 2021), in other words, they should be something made public after the date of training data (in the case of ChatGPT, something on October 2021 or later) or it can be shown that the training of the specific generative AI tool has not used data that could have brought into doubts to the quality of replication.

What Needs to be Done: Develop a series of prompts (this is the domain of prompt engineering) to recreate the target work, as close as possible, and the prompts used should require as little domain expertise as possible. It is acceptable that this replication is a combination of generative AI output and human modifications/synthesis of such output, through an iterative process.

What we will publish:

• The citation to the original work
• The replicated work in its entirety, the replicated work could be a combination of generated AI content and human modification, if any (ideally, as little human modification as possible)
• The process (e.g., series of prompts) used to complete the replication, and explanation on why this specific process is used
• Brief comments from the original author (if they wish)
• In appendix, we will publish all intermediate exploration in completing this task

To promote such research, CNS will create a space “Generative AI Infused Research” to publish these papers on a regular basis. The review process would involve one reviewer and an editor. We expect to have one round of revision for the papers that are accepted.

For questions, please reach out to one of us (minding@psu.edu or grewalr@unc.edu).

[1] For example, see https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/, accessed June 2023.