Today, the power of online reviews on consumers' purchasing decisions can no longer be ignored. There is plenty of empirical evidence to support the positive relationship between online review ratings and a business's bottom line.
Not all online reviews carry the same effect, however. Those reviews that are exposed to more internet users will have a bigger impact on a business than the ones that are barely read by others. Then, the question arises: Is it possible for a hotel manager to strategically "promote" certain online reviews so they could be seen by more future consumers?
The answer is yes. It is possible if the manager knows how a retail or online review website works. Generally speaking, all of those websites allow consumers to vote on the helpfulness of an online review and tend to feature the reviews with more helpfulness votes on the front page.
The reviews listed on the first page will have a better chance of being "discovered" and read by others. Meanwhile, managers are usually allowed to reply to an online review with one manager response.
This can be done if the manager can identify two things:
- The important factors that contribute to the helpfulness of online reviews.
- How businesses may use "manager response" to influence the helpfulness of online reviews or to moderate the impact of those important influential factors.
With that in mind, I recently conducted an empirical analysis with another key investigator, Karen Xie at University of Denver. In this study, we drew our conclusions based on a linear regression model with 56,284 consumer reviewers and 10,797 manager responses from 1,405 hotels on TripAdvisor.com.
While the detailed report can be found in International Journal of Contemporary Hospitality Management, I am going to highlight the study's important findings and business implications as the following:
Important influential factors: A proposed model
In our original model for statistical analysis, we expected the following five factors to contribute to the helpfulness of online reviews:
- Quantitative review characteristics (i.e., star ratings from 1 to 5)
- Qualitative review characteristics, including number of words and number of sentences in a review
- Reviewers' demographic background, including sex and age
- Reviewer experience, including status (e.g., a senior reviewer vs. an inactive member), length of membership and number of cities visited
- Manager response (with or without)
Additionally, two moderating effects of manager response were tested — one was on the relationship between the quantitative review characteristics and the helpfulness of online reviews, and the other was on the relationship between reviewer experience and the helpfulness of online reviews.
The research findings
We identified the following as the key influential factors that contribute to the helpfulness of online reviews:
- Reviews with lower ratings
- Reviews with fewer sentences
- Reviews written by male reviewers
- Reviews written by reviewers with higher status
- Reviews written by reviewers with longer membership
- Reviews written by those who visited more cities in the past
- Reviews with a manager response
Implications for hotels and possibly other service businesses
- Manager response is critical in general. It moderates the relationship between reviewer experience and the helpfulness of reviews in terms of reviewer status and length of membership.
- Managers may respond to selective positive reviews to reinforce the positive reputation for the business.
- Managers may respond to selective negative reviews with details of how they addressed the service failure issues, allowing potential consumers to form realistic expectations of the business.
- Because "who" writes a review matters, managers may need to pay special attention to those reviews with fewer sentences, written by male reviewers or those reviewers with higher status and/or longer membership.
- Because reviews with fewer sentences tend to be voted more helpful, managers should write a concise response just to get the key points across.
Implications for the webmasters on online review or retail websites
Webmasters may refer to our regression model and further test their existing algorithms. Then, we hope our model can help them do a better job in "predicting" those reviews with high potential of becoming "helpful" ones as soon as a product/service receives a consumer review.
If they have the power to predict the helpful reviews, webmasters can inform the business owners/managers with an auto-reminder so those reviews with a high potential of being voted helpful will be seen by the business owners/managers.
Then, the business owners/managers may decide how they want to deal with those helpful reviews. That way, the website will be able to provide more valuable service to their business partners as well as the consumers who are seeking "useful" reviews online.
Both Karen and I are hoping you will find our results and conclusions meaningful. In the end, we would like to encourage you to leave us some feedback and suggestions. Or let us know about any research or practical questions in your mind. If applicable, we would be glad to use the research questions you ask here in our future studies.