Online feedback: How the wrong system distorts the picture
In his studies, Ben Greiner, head of the Institute Markets and Strategy and the Competence Center for Experimental Research at WU, investigates which rules in online feedback systems lead to a bias in ratings.
Customer reviews are an important basis for purchasing decisions, especially for online shoppers. Hardly any online hotel bookings are made without first scrolling through all the reviews. WU professor Ben Greiner is investigating which rules in online feedback systems lead to a distortion of reviews. According to Greiner, this is not only detrimental to customers, who have less reliable information, but also to companies, which customers trust less due to a lack of information.
Whether booking a hotel or making a purchase, online reviews from others are an important decision-making tool for many people. At the same time, public reviews generate important incentives for companies to offer better products and services. However, this requires customer feedback to be honest and informative. WU professor Ben Greiner used data analysis of existing online marketplaces such as eBay and an experimental laboratory to investigate which online feedback systems best meet this requirement and discovered - depending on the feedback system - a variety of social behavior patterns that are influenced by the respective feedback rules.
"Revenge feedback" for negative reviews
On eBay, one of the market platforms on which buyers and sellers rate each other, the research clearly showed the strong reciprocity in feedback. Positive feedback from buyers was almost always rewarded with positive feedback from sellers. Negative feedback from the buyer was almost always followed by negative revenge feedback. As a result, many people who had had a bad experience did not give any feedback at all for fear of receiving bad feedback themselves. This distorted the ratings overall: 98 percent of feedback on eBay is positive. "This means that there is relatively little information in the feedback collected," says Greiner, "Our laboratory studies show that this is due to the openness of the system. If you change the rules and only publish feedback if both transaction partners have given feedback, this leads to more negative, honest feedback, but unfortunately also to less participation in the feedback system. If you only allow one-sided feedback, it also works better."
Lazy compromise
It also showed that the option to withdraw feedback can also have unwanted effects. Ideally, this option should give both trading partners the chance to resolve a conflict amicably. However, the studies revealed contradictory patterns of behavior: for example, sellers who expected negative feedback themselves only gave buyers negative feedback in order to persuade them to withdraw their feedback later. "This result is again distorted feedback. The informative value and therefore the effectiveness of the feedback system are significantly reduced," says Greiner. Overall, it can be seen that the specific rules of a feedback system determine the strategic incentives to which buyers and sellers are exposed when giving their feedback. An effective "design" of feedback systems takes these effects into account.
About the study
For his research, Ben Greiner analyzed field data from eBay and other market platforms and data from laboratory experiments. "We can't observe many details in the field. For this reason, we recreate the markets in the computer lab and let test subjects make decisions in the roles of buyers and sellers. The profits that the participants generate in the laboratory experiment are paid out to them as cash. This makes these decisions real. As in a 'wind tunnel', we can then change the rules of the feedback systems and observe the reactions of the buyers and sellers, explains Greiner.
