Use and Influence of Large Language Models for Optimizing Product Texts in E-Commerce
Ann-Cathrin Nordhoff () and
Darius Zumstein ()
Additional contact information
Ann-Cathrin Nordhoff: Migros Fachmarkt AG
Darius Zumstein: University of Applied Sciences and Arts Northwestern Switzerland
A chapter in Digital Management and Artificial Intelligence, 2025, pp 399-411 from Springer
Abstract:
Abstract This contribution investigates the impact of disclosing AI-generated text on consumer purchase intent and their perception of text quality in the e-commerce sector. As large language models (LLMs) become more prevalent in commercial text production, it is crucial to understand how consumers react to AI-generated content and the importance of transparency regarding its origin. The research results reveal that while disclosing the AI origin of texts does not significantly reduce purchase intention overall, consumer responses vary depending on their initial perceptions of content quality. This suggests that the mere knowledge of AI involvement does not necessarily influence consumer decisions negatively, especially when the perceived quality is high. Furthermore, AI-generated texts were generally perceived as inferior in quality compared to those written by humans, particularly in terms of emotional appeal and authenticity, underscoring the need for robust strategies in developing and managing AI-generated product descriptions. Effective quality assurance of such content is essential, as perceived quality is a more decisive factor in purchasing than the awareness of the text’s AI origin. This is supported by data showing a significant correlation between perceived quality and individual attributes such as usefulness, relevance, completeness, understandability, emotional content, text length, and structure of product page. Expert interviews conducted as part of the research project emphasized the necessity of tailoring texts to the specific needs of customers, providers, and products. High-quality texts should contain relevant information, clearly highlight product benefits, and be structured in a way that enhances readability and engagement. While keyword integration and ongoing SEO optimization remain important, the study’s focus on consumer perception revealed that neither AI-generated nor human-written texts were consistently recognized as such by consumers, and familiarity with LLMs did not predict better identification of AI texts. Additionally, the research showed that the impact of AI disclosure on consumer trust varies; some consumers are indifferent, while others may become skeptical if the perceived quality is low, highlighting that content quality is more crucial than transparency alone. In conclusion, this contribution demonstrates that transparency about AI usage alone is not sufficient to affect purchase decisions if the quality of the information is perceived as inadequate. Practically, the findings suggest that e-commerce businesses should concentrate on improving the quality of their AI-generated product texts and consider strategic transparency based on customer segments. This approach could foster trust and enhance customer satisfaction, while the strategic use of AI can improve the overall effectiveness of product texts.
Keywords: E-commerce; large language models; artificial intelligence; product text; product detail page; search engine optimization; ChatGPT; purchase intention (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:
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:spr:prbchp:978-3-031-88052-0_33
Ordering information: This item can be ordered from
http://www.springer.com/9783031880520
DOI: 10.1007/978-3-031-88052-0_33
Access Statistics for this chapter
More chapters in Springer Proceedings in Business and Economics from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().