Leveraging Machine learning and generative AI for content Engagement: An Exploration of drivers for the success of YouTube videos

Arindra Nath Mishra (), Pooja Sengupta, Baidyanath Biswas, Ajay Kumar () and Kristof Coussement
Additional contact information
Arindra Nath Mishra: MDI - Management Development Institute
Pooja Sengupta: University of Auckland [Auckland]
Baidyanath Biswas: Trinity College Dublin
Ajay Kumar: EM - EMLyon Business School
Kristof Coussement: Université de Lille

Post-Print from HAL

Abstract: Digital content creation has exploded in the last decade offering immense opportunities for brands and content creators. However, more research is needed on textual and aural content for determining video success using video analytics. Yet, data collection and analysis in this research context are labor-intensive. This study leveraged Generative AI (GenAI) models to automatically extract video transcripts and extract relevant metrics. We examined over 1055 YouTube videos released between 2021 and 2023 across three popular smartphones. We extracted semantic metrics from the transcript and comments to build models to explore the drivers of video success. We compared various GenAI-based measures and compared them to traditional methods. The results from this study confirm the superior performance of GPT4 over the benchmarks. The study's theoretical contributions to the field of video-based content management and the managerial implications for practitioners in the field of video analytics are discussed.

Keywords: Youtube video; ChatGPT; Large language models; Construal level theory; Customer engagement theory; Natural language processing; Video analytics (search for similar items in EconPapers)
Date: 2025-04-01
References: Add references at CitEc
Citations:

Published in Journal of Business Research, 2025, 193, pp.115330. ⟨10.1016/j.jbusres.2025.115330⟩

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:journl:hal-05531904

DOI: 10.1016/j.jbusres.2025.115330

Access Statistics for this paper

More papers in Post-Print from HAL
Bibliographic data for series maintained by CCSD ().