HOW TO USE GENAI AS A RATER IN MARKETING: A COMPREHENSIVE GENAI-BASED CONTENT CODING FRAMEWORK

Authors : Bahri Baran Koçak
Pages : 1705-1725
Doi:10.35408/comuybd.1595049
View : 50 | Download : 86
Publication Date : 2025-07-23
Article Type : Research Paper
Abstract :With the recent developments in Generative AI (GenAI) applications, popular tools such as ChatGPT have gained potential not only for individual or commercial purposes but also for the marketing discipline. As an important process for the qualitative data analysis in marketing research, the task of coding texts is mostly performed by human experts, and this can cause loss of time and increase costs. These tools may also have the potential to perform the task of coding texts, which is mostly performed by human experts for the qualitative data analysis process in marketing research. Although there are techniques applied for coding qualitative data, no study has yet presented a systematic framework to use GenAI tools in coding tasks, especially for marketing research. To fill this research gap, the current study proposes a 6-step GenAI-based content coding framework. In the framework, firstly brand messages are collected, and the themes needed for classification are determined. Then, appropriate Gen-AI models are selected to separate brand messages according to the desired themes, prompts are prepared, classes with AI outputs are coded and finally, compatibility between coders is checked. In this respect, an application is carried out to test the proposed framework. Informational, entertaining and remunerative content strategies in consumer engagement literature were used as themes and the coding agreements between 3 Large Language Models (LLMs) and human experts were determined with Kappa statistics. According to the results, the level of agreement between Human and ChatGPT gave the best Inter-Rater Reliability (IRR) for informational and remunerative contents in the comparison of Human and AI coders, while Gemini performed better for entertaining messages. As a most important practical contributions, the framework of this study offers a useful and faster coding processes for marketing practitioners.
Keywords : GenAI, büyük dil modeli, kodlayıcılar arası uyum, pazarlama, çerçeve

ORIGINAL ARTICLE URL