Persuasive Language: Evidence from Airbnb

Posted: 10 Aug 2020

Date Written: July 10, 2020


We test consumers' response to the language (e.g., classes of words, punctuation and icons) using quasi-random modifications to Airbnb product descriptions generated by property owners over time. We focus on whether language that is informative or non-informative about properties impacts the hazard that Airbnb reservations are acquired. Consistent with consumers engaged in rational search, we find evidence that consumers respond positively to factual information about the properties, a result that holds across regions and property types. We also find that non-informative language can influence consumers as much as informative, but responses to non-informative content are more context dependent. For example, consumers of properties in the U.S. respond positively to hyperbolic language (e.g, “perfection,” “enormous,” “paradise”), while consumers of properties in the U.K., Ireland and Australia do not. Additionally, these latter consumers respond quite negatively to text icons while consumers of U.S. properties do not. Both groups of consumers exhibit a similar aversion to the use of words in all capital letters. These results suggest that product information is universally valued by searching consumers, but non-informative language may influence consumers differently depending on local norms and context.

Suggested Citation

Thomas, Michael, Persuasive Language: Evidence from Airbnb (July 10, 2020). Available at SSRN:

Michael Thomas (Contact Author)

Santa Clara University ( email )

Santa Clara, CA 95053
United States

HOME PAGE: http://

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics