Product ratings and reviews should provide clarity about a product’s qualities. Unfortunately, the nearly ubiquitous star rating scale can foster assumptions; it often lacks clearly defined measurements and fails to encourage written reviews. Context is important, and in this article, we will discuss a solution to improve communication among readers for finding and sharing literature.

I love literature, and I believe it deserves better than a grading scale that elicits confusion and degrades communication among readers. In this article, we’ll imagine a better one to practice thinking and building more satisfying experiences.

The rating and review that inspired me to dig deeper into the star rating system can be paraphrased like this:

My brain stuttered when I read this comment. Had I written the review, even if I preferred other genres, I would have rated the book five stars. I expected anyone who said a book was lyrical, lovely, and haunting would feel the same; I expected that the original reviewer and I would share an understanding of what makes a book three stars versus five stars. The rating seemed at odds with the review, and I kept wondering how the original reviewer and I could be on such different pages.

The system’s vagueness allows it to be applied to every disparate thing, from books to bras to brooms. Said another way, star ratings are not optimized for any one thing.

In an attempt to prevent individual interpretations, many companies uniquely define what each star category means on their sites. However, with a widely used glyph scale, this puts an unreasonable onus on users to learn the differences between every site to ensure correct usage—the equivalent of learning a homonym with hundreds of definitions. As there are few reasons to think one five-star scale broadly differs from another, companies reusing the system should anticipate that:

Unfortunately, this creates countless little inconsistencies among user ratings that add up.

You can notice some of these inconsistencies if you scroll through ratings and reviews on a site like Goodreads, where there are a variety of interpretations of each star category. For instance:

Without speaking a common language through standardized rating definitions, readers relying on one another to discuss and discover meaningful literature becomes exceedingly difficult.

However, let’s pretend everyone using a site like Goodreads agrees on what each star category means. The problem remains that a rating still tells us nothing about what a reader liked or disliked. Without personalized context, well-defined star ratings mainly act as a filtering system. People can choose to disregard books below a specific number of stars, but even they need to learn more about the books within their read-worthy range to decide what to read. On a social literature site, users should be able to depend on others for insight.

What follows is one solution, among many, that addresses the aforementioned issues by observing three guiding principles:

The focus will be on a social literature app, but you can imagine some of these concepts applied anywhere star ratings are used. Let’s discuss each part of the solution.

The first piece of our solution primarily centers on trust, although it is not novel: readers are required to “Shelve” a book as “Read” before writing a review.

This feature is a step toward assuring readers that reviews are genuine. It also builds reviewers’ confidence that they will contribute to a credible conversation. Although users can lie to bypass the gate, honesty has more incentives on an app to find and share literature. A liar has little use for the former, and for the latter, if their intent is to gain attention for a book, they risk getting caught in a discussion that uncovers them, per the upcoming suggestions.

Simple and familiar, being mindful of people’s time, and contributing to clearness: once a reader shelves a book as “Read,” they can “Favorite” it.

Because this is a straightforward input, it requires less effort than deciphering the differences within anyone’s five-point star scale. Not favoriting a book does not indicate that you disliked it, and that is not the purpose. Favoriting tells people this is a noteworthy book for you, which may inspire them to learn why, read reviews, and interact with others. The number of times a book is favorited can be tallied to rank it in lists and garner extra attention.

In addition, vastly improving on our principle of clarity, once readers shelve a book as “Read,” the app also prompts them to mention what they enjoyed.

Respecting a reader’s time and developing a common language for users, the prompt provides a list of predefined answers to choose from. The options are mostly based on conventional literary characteristics, such as “Fast-paced plot,” “Lyrical language,” “Quirky characters,” and dozens of others.

Every quality a reader chooses gets added to traits others have chosen. On a book’s overview page, the selected qualities are ranked in descending order, equipping prospective readers with a clearer sense of if they might like a text based on top traits. The number of qualities a reviewer can choose is limited to encourage thoughtful selections and discourage abuse by selecting every trait on the list.

Similarly, there could also be a “Wished” list that follows the same structure as the “Enjoyed” list. “Wished” would create opportunities for people to mention what else they would have liked from a book, and the collective traits of reviewers could further assist in someone’s decision to read a work.

Every feature mentioned so far is enhanced by the final piece of our solution: written reviews. Allowing users to explain their thoughts, such as why they chose the qualities they enjoyed, gives potential readers a deeper understanding of the book they are considering.

In our solution, users are not merely given a blank text box and asked to write a review. The users are prompted to share their thoughts and receive suggestions to hone their feedback. The suggestions range dynamically, depending on a reader’s earlier choices. If they favorited a book, the prompt might ask why. If they chose the “Well-developed characters” option from the Enjoyed list, the prompt might ask how the characters are well developed. The prompt might also nudge users to read other people’s reviews for ideas.

Finally, these features require regular usage to benefit readers. Growing an active community around them can be accomplished by building healthy communal habits, which hinge on voices having the capacity to be heard. Thankfully, one of the oldest features of the Internet can do a lot of the heavy lifting to solve this: commenting. Many sites offer the ability to comment on reviews, but several also employ a “Like” feature — the ability to press a button that “Likes” a review or comment — and liking comments can weaken the voices of a community.

If we encourage people to not solely judge a book by its cover, we should extend that advice to its star ratings.

Let’s look beneath the surface of things and use our hearts — just as Antoine de Saint-Exupéry’s eponymous Little Prince had to learn to do — to discover meaningful new territories in books and elsewhere. A good place to start would be reading or rereading that story, marveling at the stars dotting its illustrated skies, and writing a review to share what we each found buried within the pages.

While the recommendations throughout this article are focused on how a company can change its rating and review system, companies can be slow to change. Fortunately, there are steps readers can take to be more thoughtful when reviewing the literature. A simple process today might look like this:

You can use variations of this process to review other products, too. Remember that the most important part is that we use our words. This helps reduce confusion that might come from a lone star rating, and it helps us build stronger connections.