Review Systems in Online Marketplaces

In the digital era, online marketplaces have become ubiquitous, connecting consumers with products and services with unprecedented ease. At the heart of these marketplaces lies the review system — a critical tool for building trust, ensuring quality, and guiding consumer decisions. However, as the digital marketplace landscape evolves and the traditional mechanisms of these review systems are encountering new challenges and limitations. In this segment, we explore the nuances of review systems, the emerging challenges they face, and the innovative solutions and strategies that are reshaping the way online marketplaces harness the power of feedback.

What are marketplace?

A marketplace can be broken down into three points

1. Platform for Reducing Friction: Marketplaces, using examples like Airbnb and Uber, as platforms that remove friction in finding and booking services. Unlike traditional views where marketplaces are seen as selling products or services directly, Johari emphasizes that they actually sell the convenience of reduced transaction costs and simplified access to services.
2. Meeting Point of Supply and Demand: Marketplaces are described as locations, often online, where supply meets demand. Providers (like hosts on Airbnb or drivers on Uber) offer their services, and customers seek these services. The marketplace facilitates these transactions by bringing both parties together.
3. Beyond Just Transactions: Marketplaces are not just about facilitating transactions. They also play a crucial role in building trust between participants, ensuring quality of service, and managing the overall user experience on both sides of the market (buyers and sellers).

Strategies in Review System

In the intricate ecosystem of online marketplaces, review systems stand as a critical pillar in building trust and ensuring quality. Most online marketplaces uses traditional approaches and strategies to gather and display user feedback, relying on simple average star ratings and written reviews to inform potential customers.

Here are a list of breakdown of the strategies used for a review system in most online marketplaces

1. Incentive Structures: Feedback systems are designed based on various incentive structures to encourage honest and high-quality transactions. Some systems reward positive behavior (like prompt shipping or accurate product descriptions) while others penalize negative behavior (such as late deliveries or misleading product information).
2. Anonymous vs. Non-anonymous Feedback: Some systems allow anonymous feedback, where users can rate each other without revealing their identities. Others require non-anonymous feedback, where the identities of the reviewers are known. This difference affects the behavior of users, as anonymity can sometimes lead to more honest but also more harsh feedback.
3. Unilateral vs. Bilateral Feedback: In unilateral systems, only one party (usually the buyer) can leave feedback, while in bilateral systems, both parties (buyer and seller) provide feedback. Bilateral systems can create a more balanced view but might also lead to retaliation concerns.
4. Quantitative vs. Qualitative Feedback: Some platforms rely on quantitative feedback (like a 5-star rating system), while others use qualitative feedback (written reviews). The choice between these affects the kind of information available to users. Quantitative feedback is easier to aggregate and compare, while qualitative feedback provides more detailed insights.
5. Static vs. Dynamic Systems: Some feedback systems are static, with fixed rules and formats. Others are dynamic, evolving based on user behavior and market needs. For example, eBay's feedback system has undergone several changes over the years to address issues like feedback inflation and retaliation.
6. Open vs. Controlled Feedback: In open systems, all feedback is visible to everyone. In controlled systems, the platform may filter or moderate feedback, showing only what it deems most useful or relevant.
7. Direct vs. Indirect Feedback Mechanisms: Direct feedback involves explicit ratings or reviews provided by users. Indirect feedback mechanisms might include more subtle indicators of trustworthiness or quality, like the number of transactions completed or the speed of response.

Challenges in Rating Systems

Online marketplaces rely heavily on review systems to foster trust, guide purchasing decisions, and maintain quality control. However, these systems face several challenges that can undermine their effectiveness and the overall marketplace integrity

Rating inflation

A prevalent issue in many marketplaces is rating inflation. Over time, average ratings tend to creep upwards, not necessarily due to an overall improvement in quality or service but often due to the psychological and social dynamics at play. Users may feel compelled to leave higher ratings to avoid conflict or because the norm shifts — what once was a 'four-star' service gradually becomes 'five-star' in user perception. This inflation dilutes the effectiveness of the rating system, making it harder for consumers to discern truly exceptional services from the average.

Possible solutions

One method is to redefine rating scales. Instead of the generic poor-to-excellent scale, incorporating specific prompts such as "Exceeded Expectations" can encourage users to think more critically about their ratings. Another approach is comparative rating, where users are asked to compare their current experience with previous high-rated experiences. This method grounds the rating in a personal context, making it more reflective of the actual service quality.

Another innovative approach in enhancing the integrity of review systems is the implementation of a double-blind review process (e.g Airbnb). This method ensures that neither party can see the other's review until they have submitted their own. This approach aims to foster honesty and prevent bias in the review process.

Understanding the Sound of Silence

The absence of a review, or the "sound of silence," is often an overlooked aspect of review systems. Not all users who have a negative experience leave a review; many choose to remain silent. This non-response can be a significant indicator of dissatisfaction. Marketplaces must consider these silent signals and find ways to incorporate them into their overall assessment of a service or product. It can be encouraging feedback proactively through rewards, simplifying the review process, implement mechanism in following up abandon reviews or incorporating indirect feedback (e.g social media, forums).

Distributional Fairness in Ratings

Averaging ratings can be unfair, especially for new entrants in the marketplace. For instance, a single negative review can disproportionately affect a new participant compared to an established one with hundreds of reviews. This disparity can stifle new entries and reduce marketplace dynamism. To address this, some platforms are considering Bayesian approaches that incorporate a 'prior belief' into the rating. Another way to address this is to implemented weighted rating system (less impact for first few reviewers) or time-weighted ratings (recent reviews weighted more than past reviews) or implement a threshold before displaying rating averages. With these methods, it can balance out the impact of early negative reviews, giving new participants a fairer chance to establish themselves.

Conclusion

Review systems are more than just a feature; they are a fundamental aspect of marketplace integrity and user trust. Rethinking these systems involves understanding the complex dynamics behind ratings, innovating beyond traditional scales, recognizing the importance of non-responses, and ensuring fairness for all participants. By doing so, marketplaces can enhance the accuracy and reliability of their review systems, fostering a healthier and more competitive environment.