Ann Kronrod, University of Massachusetts Lowell – What Can We Learn from Exploring The Language of Product Reviews

Do product reviews matter?

Ann Kronrod, associate professor of marketing at the University of Massachusetts Lowell, reads up to find out.

Ann Kronrod is an Associate Professor in the Department of Marketing, Entrepreneurship and Innovation. She earned her Ph.D. in Marketing and Cognitive Science of Language from Tel Aviv University, and later completed her education as a Postdoctoral Researcher at MIT, Sloan School of Management. Prior to joining UMass Lowell, Ann Kronrod was an Assistant Professor at Michigan State University, and then Visiting Assistant Professor at Northeastern University and at Boston University. Ann Kronrod is a marketing researcher with extensive background in linguistics. Her research interests span a wide variety of subjects that can be categorized as marketing communication, consumer behavior, word-of-mouth and pro-social marketing. She often integrates her knowledge of linguistics in her research.

What Can We Learn from Exploring The Language of Product Reviews

For the past 20 years, the reviews platform YELP has been providing consumers with an indispensable opportunity to share their experiences at restaurants, medical facilities, and so on. Learning from hundreds of consumers improves our judgment about products and services, but with the recent proliferation of AI-generated reviews, fake reviews are becoming an epidemic. Sadly, we, humans, are awful at telling truth from a lie. We are correct about 51%-54% of the time, which is basically the rate of flipping a coin.

I’ve been investigating Language in Marketing for the past 20 years and together with my colleagues[1] I developed an algorithm that detects fake reviews by the language they use. We found that if a consumer tried the product, their vocabulary comes from episodic memory of their experience and will contain more concrete and unique words, and sentences that relate to the what–where–when (WWW) of the experience. By contrast, if you are just making up a fake review, you simply won’t have those words in your vocabulary. Instead, you will be using words from your semantic memory: memory of general facts faintly related to the experience. So, your language will be more abstract and general and it will be unrelated to time and place. We ended up with an algorithm that detects fake reviews by their language and can be used on any set of reviews.

This way, our research helps platforms like YELP overcome these negative developments threatening consumer-firm trust and customer experience.

[1] Kronrod, Gordeliy and Lee 2023

Read More:
[now publishers] – Language Research in Marketing

Share

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *