With Reviews, Simpler is Not Always Better
If you are trying to plan an event like a night out, a work trip, or a vacation, you most likely utilize reviews of some kind. Whether it’s Yelp, Google, or another review website, typically these reviews utilize the same system: a simple 1-to-5-star rating. On the one hand, the system checks a lot of boxes: simple to use, easy-to-understand the results, and it can apply to all sorts of businesses like restaurants, hotels, and airlines. So, it’s not surprising that so many companies and people rely on this system to inform consumption choices, but is this the most efficient way to display and inform consumer opinions?
When looking for products, according to a 2014 PowerReviews study, 95% of customers will read reviews before finalizing their purchases. This year, Bright Local puts the number at 98%. Not only do almost all consumers read reviews before purchasing an item or making a reservation, but they also hold extremely high trust in these reviews, with Yelp saying 91% of people trust online reviews as much as personal recommendations. While some of these results could be viewed as a little self-serving, it feels safe to say that a majority of consumers based their purchasing decisions at least somewhat on the reviews left for products or places.
On the surface, this seems fine and the straightforward 1-5 star rating feels like a good match. The challenge, however, is that simple is not always best and also that the review system has created a system of grade inflation that would make Harvard blush. In a well calibrated system, there would be something close to a bell curve, where most products receive a middle rating and there are a diminishing number of reviews as you move out to the very positive or very negative. In the ideal situation, the average rating (on a 1-5 scale) would be close to 3.0.
However, The Next Web in 2016, looked at Amazon ratings (in a very well titled piece) and found that the average rating on Amazon is 4.4 overall and increased to 4.73 when companies incentivize reviews. The study points out that while the 0.33 rating difference between the incentivized and non-incentivized might seem small, an average product with a 4.4 star rating jumps from the 54th percentile to the 94th percentile once you factor in the ratings bump.
Similarly, Rising Star Reviews reported that on Yelp the average review is 3.78. While this is closer to the 3.0 ideal average than what’s found on Amazon, the breakdown of the results shows that the reviews have become extremely skewed, with a majority of reviewers giving 5 stars to restaurants and more than two-thirds (68%) giving a 4 star review or higher.
Both restaurants and sellers on Amazon have become very aware of these numbers and trends and, due to this reliance on reviews, they know that negative ratings can have detrimental effects on businesses. In a few studies, about half of consumers are unlikely to buy a product or service with below a 4.0 star rating. At the same time, there is a mistrust of products that don’t have any bad reviews. This has turned review consulting into a cottage industry with consultants happy to walk you through the best way to get the right kind of reviews. Similar to college rankings the inputs are being manipulated by most to get the highest ratings possible.
Besides the fact that the process has been corrupted, the reliance on simplicity also turns every product or experience into one big bland lump of vanilla. Even though every consumer approaches decision-making differently, we all have our own personal biases, and each purchase or decision is being made for different reasons, we get the same simple prompt: “Please rate this product on a scale of 1 to 5.”
Context is everything, yet we are not able to give any additional perspective to our feedback. Similarly, from the consumer standpoint, potential buyers get just a few pieces of information and the ability to narrow down to your desired level of feedback is either non-existent or would take hours of time to figure out. For example, let's say a couple is going out for an anniversary dinner and are looking for a fine dining experience, but ends up at a Wendy's. Should their 1-star rating be on equal footing with the family driving on vacation who needed a quick lunch and gave it 5 stars?
Another example is someone traveling for work who is staying a couple nights in a city in order to hold a handful of meetings while in town. Their review of the hotel is considered equal to the couple looking for a romantic weekend away or the family on a trip for a long weekend. The exact same hotel can garner 5 stars for the business traveler, 3 from the family, and 1 from the couple, but does an average score of 2.7 help anyone in this situation?
The truth is that the context surrounding the purchasing decisions is just as important as the rating itself. Most businesses have a specific audience that they cater to and reviews from those outside this audience are much less relevant. Yet the review has become omnipresent. Do we really need reviews for products that never change? Q-Tips are a fairly simple example. Why would you need to read reviews on Q-Tips? Maybe for a non-brand product of cotton swabs, but search around Amazon and you can find multiple examples of different reviews for the same product. Looking at Amazon one 500 pack of Q-Tips boasts an impressive 4.8-star rating. Is that helpful?
As consumers, we should want the most accurate information to inform our purchasing decisions. However, businesses should also seek for more accurate reviews as it would prove much easier to tailor their practices to the audiences which they seek to target. Perhaps this simple 1 to 5 star rating system is not working any longer and we need to start looking for other options which will better serve consumers in locating products and services that will satisfy the exact niche for which they are searching.
The bottom line is that, as technology improves, so should our access to better information. This could take many forms, but one option would be giving purchasers and users more options on which to provide feedback. Some stores have implemented basic versions of this approach, such as REI, which asks for feedback on additional categories such as fit and durability. This is a good start, but what really needs to happen is that we need to move into more of a conversational dialogue than just pick one single numerical rating out of five.
At the same time, the ease of feedback needs to be at a similar level when it comes to the ease of use for the reviewer. Very few people will spend 10 minutes writing down every detail. However, a conversation starts with an understanding of why a consumer purchased that particular product, and then factoring that information into the feedback given. If we connect the why with the review, then, on the other end of the feedback loop, the potential customer can put in their parameters and see the reviews of people who match their specific situation.
Using the hotel situation from above, if a business traveler is able to share information about their trip and their review is anchored to the fact that this person just needed a clean place to stay in a specific area, a different business traveler could easily find that feedback. Similarly, the business traveler could easily be asked if other types of travelers would feel the same.
The technology is there to move past the most simplistic feedback possible, so now feels like the right time to start using the tools at our disposal.