Bayes Rule, Yelp Reviews, and Customer Satisfaction
Unfortunately, what we see or are told is not always what we get. While this applies to many aspects of life, I have found this to be especially true when I order food from restaurants. When I am somewhere where I do not have experience with the restaurant options, I have no option but to turn to people who are knowledgeable about the food places around me. When not with people from the area, the next step is to consult people on the internet through popular websites such as Google and Yelp. When in a new place, many people make decisions based upon “herd behavior” or the wisdom of the masses. In the context of the restaurant example, people are more likely to try the restaurant with many 5-star reviews than one with many 2-star reviews.
However, according to a blog titled “Are Yelp reviews reliable?”, nearly 20% of online reviews are fake and many businesses can pay Yelp to put their good reviews in front of the customer first. The origin of the reviews being misleading stems from review manipulation where companies can fake reviews to damage competing businesses, from customers using reviews to their advantage to get discounts, and from companies can leaving positive reviews of themselves.
This article is not meant to discount the validity of all restaurant reviews; however, it is meant to show that good restaurant reviews are not a guarantee of overall satisfaction. This can be illustrated by Bayes rule which calculates the probability of an event based on prior knowledge about the event. In our example, this would be the probability you are satisfied with the restaurant in correlation to the restaurant’s overall reviews and satisfaction. Bayes theorem is that there is no guarantee that the event happens (the customer enjoys their dining experience); but Bayes theorem seeks to explain the probability that the event happens based on a condition (a good review of the restaurant they review on Yelp) surrounding the event. While it is impossible to calculate the actual probability a restaurant is good based on a yelp review due to not knowing the amount of fake reviews, subjective satisfaction ratings due to personal preferences, and other unaccountable variables, the logic of Bayes rule remains consistent that certain conditions can change the probability of an event based upon conditions revealed before engaging in the event. In our example, if a customer chooses a higher rated restaurant, the probability is altered if they chose the restaurant at random, and they are more likely to enjoy the restaurant as many other people who reviewed the restaurant did as well.
https://blog.reputationx.com/are-online-reviews-reliable