aceModerate evidence
More
Expert and step-by-step prompts beat basic ones on hotel and tourism tasks, experiments find
Study title: Artificial Intelligence prompt engineering in tourism and hospitality: a task-technology fit theory on performance, complexity, and user literacy
- Evidence type
- Controlled or validated comparison
- Sources
- 1 sourcePeer-reviewed
- Published
- 1 min read
- Truvace record
- TRV-2026-2649cite or check it
Study title: Artificial Intelligence prompt engineering in tourism and hospitality: a task-technology fit theory on performance, complexity, and user literacy
Two experiments tested how different ways of prompting generative AI perform on marketing and management tasks in tourism and hospitality. Structured reasoning prompts, particularly expert prompting and chain-of-thought, significantly outperformed basic prompts.
Read the study abstract
Despite the rapid integration of generative artificial intelligence into tourism and hospitality operations, there is a critical gap in the empirical framework for evaluating how different artificial intelligence prompt engineering strategies perform across diverse marketing and management tasks, and examining how user characteristics shape these outcomes. This study addresses this gap through task-technology fit theory across two complementary experiments. Study one employed a 16 × 2 factorial design, assessing…
The gain
In two experiments on tourism and hospitality marketing and management tasks, structured reasoning prompts, especially expert prompting and chain-of-thought, significantly outperformed basic prompts, and user AI literacy did not significantly change how well prompts worked.
In brief
Task complexity limited AI performance across every prompting strategy tested. Prompt quality helped, but harder tasks still produced weaker results.
The authors also tested whether a user's AI literacy, low or high, changed how well prompts worked. It did not significantly, which they read as a sign that modern AI interfaces open access to different user groups. They suggest organisations build prompt libraries and tier their AI deployment.
Main points
- Study one used a 16 × 2 factorial design, comparing 16 prompt types across simple and complex task conditions.
- Study two used a mixed-factorial within-subjects design and added AI literacy (low vs. high) as a moderator.
- Expert prompting and chain-of-thought significantly outperformed basic prompts.
- Task complexity constrained AI performance consistently across prompting strategies.
- AI literacy did not significantly moderate prompt effectiveness.
- The authors recommend organisational prompt libraries and tiered AI deployment in tourism and hospitality.
The full story
The research applies task-technology fit theory to conversational AI in tourism and hospitality. The authors say there was little empirical evidence on how prompt strategies perform across diverse marketing and management tasks.
Study one compared 16 prompt types on simple and complex tasks. Study two had participants work within subjects and compared low and high AI literacy groups to see whether user skill shaped results.
The authors conclude that the findings extend task-technology fit theory and support shared prompt libraries and tiered deployment, where task difficulty guides which tools and prompts are used.
Sources
1 on file- Tourism Recreation Researchdoi:10.1080/02508281.2026.2727979
Talk it over
Who gets this benefit, and who was left out of the study or story?
Expert and step-by-step prompts beat basic ones on hotel and tourism tasks, experiments findhttps://truvace.com/item/artificial-intelligence-prompt-engineering-in-tourism-and-hospitality-a-task-tec
Who gets this benefit, and who was left out of the study or story?
The gain In two experiments on tourism and hospitality marketing and management tasks, structured reasoning prompts, especially expert prompting and chain-of-thought, significantly outperformed basic prompts, and user AI literacy did not significantly change how well prompts worked.
Answer in the threadHow do we know? What can this kind of evidence show, and what can’t it?
The evidence This rests on a controlled comparison, from Tourism Recreation Research.
Answer in the threadWhat does it leave open? What else would you want to know before acting on it?
Answer in the threadHave you seen AI used this way where you study, work or live? Did it go the same way?
Answer in the thread
The questions are the same on every record. The details under them are taken from this record as written.
What readers say