A Hybrid Approach to Identifying Target Audience Pain Points: A Comparison of In-Depth Interviews and AI-Assisted Prompts

This article compares a traditional deep-dive interview with AI-assistive prompting to surface how pain points of the target audience are identified.That is, latent user needs and affective attitudes must be surfaced quickly and scalably in response to the increasing demand by companies.The justification for this study lies in the high cost and time inertia associated with classical in-depth interviews, coupled with the potential applications of generative language models in speeding up the preliminary analysis of customer pains.This paper conducts an empirical comparison of resources, time consumption, and quality outputs between 20 interviews (approximately 720 minutes) and 20 AI-assisted session logs structured under the ABCDX and BDF frameworks.The novelty of this study lies in its mix-and-match method, which joins speed checks in idea generation with content analysis and attestation of AI-generated ideas from actual conversations.All that information is brought to one place-here-in detail, ordered by how often and how strongly it is mentioned.Findings almost fully converged on the pain points; over 80% of hypotheses generated by LLMs were validated during interviews.Moreover, it takes about 44 minutes less to complete the initial analysis each time after AI assistance (9 minutes per session).In-depth interviews retained their importance at the stages of validation and refinement of -disputed topics because clarifying context, including nonverbal signals, proved crucial.The optimal protocol recommends starting with an AI session for broad hypothesis generation, then proceeding to targeted in-depth interviews for confirmation and elaboration of key insights, enabling up to 50 % time savings over classical qualitative analysis without compromising result reliability.This article will be helpful to researchers in marketing and user experience, product managers, and analysts engaged in qualitative research.