This page is part of a global project to create a better online reviews system. If you want to know more or give your feedback, write at [email protected] and we’ll grab a beer ;)
People tend to leave reviews only when they are extremely satisfied or extremely disappointed, often based on exceptional events. As covered in “Why do we leave online reviews”, that’s because emotions are a significant driver for leaving a review. Research has shown that consumers are more likely to share their opinions when their experience is either very positive or very negative, especially when it deviates from their expectations formed by existing opinions $^1$. An “okay” experience is rarely reviewed.
This means that the set of reviewers is not a fair representation of all customers by default—at least organically.
To encourage “average” customers to leave reviews, companies try various incentives: Airbnb only displays the review you received after you leave one, Google incentivizes “local guides” with points that can be converted into advantages, and Yelp has an “elite squad” of users who get invited to private events. These are some of the catalysts mentioned in “Why do we leave online reviews.”
There are several reasons why people might not leave reviews, which we’ll cover in the next sections: feeling uncomfortable judging others, lack of time, uncertainty about what to say, privacy concerns, and poor timing. All of these make strong motivation even more necessary.
Another important point: no review can sometimes be as bad as a negative review. Studies $^2$ showed that people often avoid leaving a review after a poor experience. This may be due to feeling bad about it or fear of retaliation, among other reasons.
This leads to an over-representation of “extreme” reviewers, whose motivations are stronger.
<aside> 💡 Exploration
By addressing the issues in the reviewing process that we’re listing here, we can reduce the dominance of extreme opinions by making it easier for those with an “average” experience to leave a review. However, this won’t eliminate the natural tendency of humans to be driven by emotions.
An AI assistant to help users articulate their opinions and add nuances. When a reviewer writes a comment, an AI could analyze the text and offer suggestions. One potential downside is that the reviewer may feel under a certain scrutiny.
Example of a review with an AI assistant suggesting to add details
“Yes, Because, But” framework. We could guide people by asking a specific question requesting a “yes or no” answer, then ask why they feel that way, and finally ask if there was anything good (or bad) they experienced despite their overall feeling. This framework is often used in real life when someone asks for a recommendation from a friend or connection and could be applied to online reviews as well.
Example of a Yes / Because / But framework
Private comments (as used on Airbnb) can help reduce the aggressiveness and subjectivity of negative public reviews since users have the opportunity to express their frustration directly to business owners. A study showed that when people can leave private comments in addition to public reviews (as on Airbnb), they tend to leave better public reviews because they can express their frustrations privately. However, this has a downside: users might omit some negative points in their public reviews, which could be useful for potential customers to know.
Require written comments along with ratings. The impact of a new rating on the average score increases the fear of retaliation. If users have to justify their rating with detailed comments, it reduces the likelihood of arbitrary ratings. </aside>
$^1$ “Online Product Opinions: Incidence, Evaluation, and Evolution”, Moe and Schweidel, 2012
$^2$ "Vocal Minority and Silent Majority: How Do Online Ratings Reflect Population Perceptions of Quality", Gao et al., 2015
https://tally.so/embed/wkZb11?alignLeft=1&hideTitle=1&transparentBackground=1&dynamicHeight=1&page_name=extreme_opinions