If you have a strong opinion on anything that just happened to you, odds are you can find a place on the internet to leave your thoughts on a scale of one to five stars. If there’s no feedback form on Yelp, Amazon, or a company’s website, you can probably find it on Google business pages. The trickier thing might be finding a product or service that doesn’t offer user reviews.
Despite reviews coming to play such a big part in our lives, academic research has focused more on reviews for experiences like restaurants or travel, and less on important fields like healthcare, one University of Delaware researcher says.
Gang Wang, an associate professor of management information systems at the Alfred Lerner College of Business and Economics, has new research that could help customers and businesses alike understand how reviews influence some of our most critical decisions, from deciding on a surgical procedure to picking a college class.
He joined with three other researchers on the study, “The Impact of Process- Versus Outcome- Oriented Reviews on the Sales of Healthcare Services,” which has been accepted in the journal Information Systems Research, a leading journal in the field. His collaborators include Hongfei Li of the Chinese University of Hong Kong, Jing Peng of the University of Connecticut, and Xue Bai of Temple.
The idea behind the study is that people often post reviews about an experience that’s supposed to be enjoyable, like dinner at a restaurant or a cruise to the Caribbean.
But increasingly, people are also posting reviews about services that are not really supposed to be fun, like facial surgery, diets or gym programs. Unlike a visit to a deli, these decisions are more about the end goal, like looking or feeling better, than the experience along the way. They can also take months to unfold.
That made Wang and his associates wonder about how customers respond to these kinds of reviews, and if it made a difference if the evaluations focused on the process — how things went with the personal trainer not just on day 1 but on day 42 — versus a review focusing on the hoped-for outcome, like a trimmer physique.
Fortunately, they were able to find a massive trove of data to test their ideas. It came from an online cosmetic surgery site in China that connected 25 million registered users to thousands of different providers, and crucially, also provided data about the kinds of services these customers picked and when. Customers also filled out online diaries that walked readers through their experience. (Services like realself.com offer a similar experience to English-speaking users.)
To sort through all that data, ever-improving artificial intelligence technology was crucial. The researchers trained AI to analyze images in the reviews and sort them into the appropriate categories. They had to manually label 5,000 photos, but happily allowed the machine intelligence to handle the bulk of the 2 million images. Testing found the AI’s results to be reliable.
Wang had previously worked with AI to analyze text, like with Yelp reviews, but says the technology has developed extraordinarily quickly in the past few years and he sees promise in analyzing not just images, but videos. “That’s going to be very, very exciting.”
From this analysis, Wang and his cohorts found that different kinds of reviews did produce very different customer reactions, but that it depended strongly on the circumstances.
If people were signing up for a popular or simple procedure, they were more influenced by reviews that focused on outcomes. With so many people getting Botox injections, for example, they might not worry about the procedure. They’d just want to know it would get rid of wrinkles as promised.
If on the other hand, customers were in the market for a more complicated, or less popular procedure, they were more influenced by reviews that documented what they could expect from each step — how much swelling the surgery caused, how long it took to heal, and so on.
“I think the key message from this study is, it really depends on the types of the procedure or the services,” Wang said.
Both consumers and businesses can learn from the results, according to Wang.
One lesson for consumers is to avoid being trapped by simple “before and after” advertising, especially for risky and complex services, Wang said.
More positively, if reviews allow customers to familiarize themselves with the process, patient satisfaction could improve, Wang thinks. Even if they are embarking on a difficult experience, they know what to expect.
At the same time, businesses can offer options to more effectively give customers the information they need to make decisions.
The findings could also have application beyond healthcare to other outcome-focused fields like education, the study noted. For example, students could weigh in not just after a course, but while it’s going on, describing their experience with factors like workload, Wang said.
“The outcome definitely is important, but also the process, right?” he said. Pointing to healthcare, Wang added, “How I am going to go through the recovery process is very important for me to make (an) informed decision.”