Research shows even positive online reviews are a minefield for firms

Customer's online reviews of products and services are highly influential and have an immediate impact on brand value and customer buying behaviors.

According to the Pew Research Center, "82% of U.S. adults say they at least sometimes read online customer ratings or reviews before purchasing items for the first time, including 40% who say they always or almost always do so." How and whether to respond to online reviews is a critical consideration that may very well make or break a company's success.

Rensselaer Polytechnic Institute's T. Ravichandran, Ph.D., professor in the Lally School of Management, and Chaoqun Deng, Ph.D., of Baruch College's Zicklin School of Business, have created a roadmap for firms to preserve and enhance positive consumer perceptions, as well as improve negative perceptions and avoid highlighting them.

In research published last year, the team formulated strategies for firms to address negative reviews. They found that, by and large, it is best for firms to custom-tailor responses to address criticisms and explain actions taken to communicate care, concern, and attempts at remedying any problem.

In their latest work, Ravichandran and Deng dove into the more complex area of addressing positive reviews. Contrary to popular perception, they found that it is not always effective to respond to positive reviews, and, in some cases, responding may make customers feel more negatively toward the company.

"We analyzed 10 years of data from Trip Advisor, which included reviews of all the hotels in Boston and Honolulu," said Ravichandran. "We categorized a portion of the reviews using natural language processing and then used a deep learning algorithm to produce our results. We found that there is no one-size-fits-all solution."

First, the researchers categorized the reviews as either instrumental or affective. Instrumental reviews are those focused on the characteristics of a good or service. They are objective, informative, and not ambiguous. Affective reviews, on the other hand, are emotional expressions about the customer's interaction with a company and tend to be ambiguous and based upon an individual experience.

Next, the reviews were categorized as either one-sided or two-sided. One-sided reviews are either positive or negative. Two-sided reviews point out both aspects.

Based on these categorizations, Ravichandran and Deng created a roadmap for managers on how and if to respond to online reviews.

They theorized that when the positive reviews are not ambiguous and have high information value, tailoring the response to the review will be effective. In other instances, a templated response or not responding at all would be better.

"Customers see tailored responses to instrumental positive reviews as attempts at self-promotion," said Deng. "They also view it as an unnecessary interference to the information they are seeking, much like a pop-up ad may interrupt your web searches."

They found that it is best to respond with a templated thank you message to one-sided, instrumental positive reviews. For two-sided instrumental reviews, the researchers suggest addressing the criticisms and outlining the actions taken to mitigate the problem.

Two-sided, effective reviews are more complex. Since they are ambiguous, it is difficult to identify and address specific issues. Therefore, the researchers conclude that a templated response is best. A tailored response may unintentionally highlight negative experiences.

Responses to one-sided, effective, positive reviews should be tailored. For mixed reviews, it may be best not to respond at all to avoid unintentionally highlighting negative experiences.

With the large volume of reviews, many companies rely on bots to respond. However, the researchers urge caution.

"Bots should be programmed so that a review either gets flagged for response or it is addressed with a templated response," said Ravichandran. "However, the determination should be a judicious one made by senior executives."

"User-generated content has a real impact on firms, and how to handle it is a challenge," said Chanaka Edirisinghe, Ph.D., acting dean of Rensselaer's Lally School of Management. "With the proliferation of fake content, firms' ability to sort through everything is increasingly challenging. A one-size-fits-all strategy will backfire. There are always unintended consequences and understanding them allows a company to act intelligently."

The findings are published in the journal Information Systems Research.