My Uber rating is 4.8, and most of the drivers Uber assigns me have between 4.7–4.9
I would think my Uber rating is just above average. In itself, it is insignificant, but combine it with the fact that the drivers who pick me up tell me that even they don’t get rides from riders with poor ratings and that as their ratings kept on getting higher, the fill rate they have witnessed in them getting rides has been soaring up — now it is beginning to sound like something.
And once you start thinking about it, more questions come up. Does a high rating makes drivers more willing towards accepting a ride or are they indifferent towards it? (Most I have asked seem to be indifferent to it.)
Has the rating become an important factor for me? (I would like to say No, but if my rating has gone up or down by even 0.01 in the past 6 months, I think I would have noticed that 9 out of 10 times.)
I have taken thousands of ride on Uber. Now, given the fact that my rating is high even after these many rides, does that mean that Uber treats my ratings to drivers differently as compared to say a new rider? (This is a question I don’t have an answer to.)
RATINGS ARE EVERYWHERE
Right from the products you see on Amazon to the rooms you look up on Airbnb to that online guitar course you checked out this weekend, ratings have become a elemental part of our consumer lives. Consumers rely on these ratings to get over their inhibition in trying out something new (community feedback), and help themselves drive decisions faster. Businesses rely on them to weed out the bad vendors and products from good ones, ensure that the consumer experience continues to be top-notch, and countless other strategic decisions.
Ratings helped democratize the overall feedback mechanism and shifted the balance of power from the few (domain experts) to the many (endless consumers out there). We rely on IMDB ratings to decide whether we should watch a movie or not, we look at ratings of restaurants on Yelp and Zomato to decide if we should step into them, and check ratings and reviews on TripAdvisor to plan our next vacation. And it is not just consumers who are using this information for their benefit; businesses use it maybe more. Quora promotes stories that tend to get more upvotes and shares, our Medium feeds are full of stories that have thousands of claps, and if I ask you how Quora or Medium or Facebook would decide what to recommend you in your feed, ratings/upvotes/claps/likes is the answer — and in some way or the other, it is just another avatar of the 5-star rating you see on Uber.
WHY DO RATINGS MATTER?
More than anything else, ratings are a convenient, quantitative, non-intrusive and visually simplest way to help consumers take quick decisions. They help consumers feel more confident about their purchases and avoid transactions which can potentially end into a nerve-wrecking post-sales experience.
But, these same ratings that help consumers make decisions faster help businesses as well, just by reducing the decision making process at the end of consumers. There are two broad classifications in which our transactions fall — (1) Those that are driven by need (requirement-based purchases), and (2) Those that are driven by desire (impulse purchases). It is primarily the impulse purchases where businesses want us to act fast and place an order before we second guess our decision to hit that ‘Buy Now’ button. Ratings, in this case, act as a facilitator.
Such is the power ratings hold — both in the lives of consumers, as well as businesses. And with great powers comes the risk of great misuse. Are we able to use this feedback in a way that is helping us drive our decisions optimally or are some retailers and businesses exploiting this in a way that just benefits them without factoring in what’s best for the consumers? How are the ratings being monitored and governed? Can organisations be trusted to audit it themselves or would it need an external audit?
PROBLEM WITH RATINGS AND REVIEWS
Let’s circle back to Uber reviews again. Most recently, Uber has updated their app to allow riders to just skip giving feedback, and regretfully I am one of those people who exercises that skipping option most frequently. And in doing so, I am not contributing to a process that can help Uber ensure that their riders get the best drivers possible, and that the driver partners who outshine everyone else get rewarded in some way or the other. Essentially, I am doing something that I should not be doing, but since I am not obligated to do so anymore, a lot of times I end up skipping on the rating.
But, what Uber used to force riders earlier wasn’t good as well. Earlier, I needed to compulsorily rate my last ride before I could book the next one, and if I gave the driver anything but a 5-star, I was forced to explain my reasons behind doing so. Rationally, it looks like a fair attempt to understand the users in an effort to deliver a better service. But in reality, it is one more thing for me to do. As a result, to avoid the additional work, I may give each ride of mine a 5 star — whether it was deserved or not. Sure, in case of a horrendous experience, I would take the time to vent out my frustration in detail, but anything short of that, I may always dole out 5 stars out of sheer convenience. And that defeats the entire purpose of the rating system.
(Just a side note — I did use to rate most of my rides 4 stars. Some rides would get the 5 stars, and some would get lower as well. But most? They got the 4 star, since convenient! Why was I miser with doling out 5-stars? Because I don’t really subscribe to the “5-star norm” that Uber has more or less made mainstream now. I do believe a 5-star rating needs to be earned. You don’t get the exemplary grade for simply doing your job, you need to really earn it. Even today, though I can skip the ratings, the rides that are just perfect do get a 5 star ratings, and the bad ones receive bad ratings. The others, I just skip.)
THE SOLUTION -or- HOW UBER FIXED IT
So, I was on an Uber ride and during the conversation, the driver complains:
Uber messes up our ratings in quite a weird way. My rating goes down from 4.8 to 4.5 after one bad ride/complain, but then it increases with a snail like speed.
And that made me wonder what could possibly be the reason behind such a behavior. It was then that I found that it is not the driver’s cumulative average rating that is displayed on the app, or treated as the driver’s rating. It is the average rating from the last 75 trips — if I recall that number right! And it blew me away!!!
Just think about it. Driver behavior can change, his attitude can change, hell he could have been putting up a show initially and now he is on his true self. There can be n number of reasons why the experience of a rider with a driver today would vary immensely as compared to a rider with the same driver months back. And yet, if cumulative average was to be considered, his past accrued high ratings would offset any negative ratings he would be receiving today. So it makes perfect sense to factor in recency while calculating the driver’s rating. The same goes for the rider.
And this. This is something which I think Amazons and Flipkarts of the world can learn from. I may feel dissatisfied with my purchase because my experience with the installation service that came with my purchase was horrible. The customers who bought the item 6 months back may have had good experiences, but today its a different agency taking care of the setup, installation and demo, and that one part is fucking up the customer experience. As a consumer, is that not something that I MUST know of? But, if I see a 4.3 rating with thousands of ratings and reviews, I may make the purchase assuming all will go well.
Don’t you think there are multiple variables that need to be factored in when we think of what rating a customer should be seeing on the product page?
SO WHAT SHOULD YOUR BUSINESS DO?
- Have a rating mechanism that benefits every single party — the consumer, the merchants, as well as the business itself.
- Remember that just like your customers, ratings are also fickle. What the rating of a product should be needs to be dynamic and factor in different parameters that can change the rating of a product at any time.
- Ensure that the ratings are not being manipulated and that they reflect the true sentiments of the consumers. (Say, when consumers — in a frenzy — start uninstalling and giving a product bad ratings on app stores because of reasons unrelated to the product, one needs to weed those out. For examples, consumers confusing Snapdeal with Snapchat started uninstalling and downvoting Snapdeal. From what I recall, Google was quick to weed those out the play store.)
As is quite evident from the countless ways in which we interact with rating mechanisms on a daily basis, they form an integral part of our online shopping behavior. Let us use them in a way that best benefits our consumers and the business.