Like the “marvelous spiral” that appears throughout the natural world, the J-curve—a pattern where a curve initially falls, then rises to higher than the starting point—is a natural data pattern that seems to be everywhere one looks:
- In private equity, funds tend to deliver negative returns in early years, but investment gains in later years
- In political science, the J-Curve is used to describe the cause of revolutions based on failed expectations
- In medicine, it is used to map the symptoms of some diseases against the chance of developing that disease
The natural distribution of star ratings forms a J-curve, where we see something like this:
As you can see, ratings tend to be positive—the average rating by client is 4.3 out of 5 stars, and 82% of ratings are 4 stars and above, but on the negative side of the spectrum you’ll find slightly more 1-star than 2-star ratings.
In some industries, we see a wider distribution of ratings (and a shallower climb). For instance, here’s the distribution for Telecom:
Companies with lower Net Promoter Scores (as is typically the case in the telecom industry) can expect a wider array of ratings, and lower average ratings overall. Fortunately, so many companies are now encouraging open and honest feedback, displaying this user-generated content on-site, and doing a great job analyzing negative reviews for ways to do better by their customers.
But even companies that have more normal J-curves, and fewer negative reviews, can mine for improvement ideas by looking at the positive end of the spectrum—20% of all 3 and 4 star reviews contain product suggestions.
Causes of the curve
There are a few explanations for why we see the J-curve in ratings. Over time, ratings may be more positive because products with many negative reviews are being pulled from the shelves, leaving better products that get better reviews. We also may be seeing a positive feedback loop—people are buying positively-reviewed products more than negatively-reviewed products, then returning to contribute reviews that will be similarly positive. As commenter Andrew Ward put it, “In this situation the accurate representation of a product and its relative price should normally drive a J Curve.” What we are certain of is that this pattern exists across every one of our clients’ verticals.
Good feelings, high ratings?
Looking at the data, we see higher 5-star ratios in product categories that are strongly associated with positive feelings and experiences. Flowers/Gifts have the highest 5-star ratio (75%), followed by Jewelry (70%) and Pets (68.6%). Think about it—we associate many of these with emotionally-satisfying moments, like weddings, birthdays, and anniversaries. Studies like this one have shown that, “pets can serve as important sources of social support, providing many positive psychological and physical benefits for their owners,” and this probably explains the higher 5-star ratio in the Pets product category.
Moving down the ratings totem, the Travel category (which includes restaurants in this case) has the highest 3- and 4-star distribution (13.67% and 34.59%, respectively). This is to be expected—many of our experiences when traveling and dining are mediocre, and we’re not always traveling and visiting restaurants for leisure. We tend to play favorites with restaurants, and there’s a good chance of being underwhelmed by an experience that’s “Just OK.” If we were able to break down reviews into work-related versus vacation/leisure, I’m confident that we would see higher ratings on the latter category.
So why exactly does the J-curve exist? We could speculate that it’s due to the positive tendencies of mankind, or rather, that it’s simply a positive feedback loop reinforced by businesses that pull shabby products from their shelves. Either way, the phenomenon is yet another reminder that your brand should never fear the feedback.