Hi @nityesh , I think that choosing the winner based on pageviews has measure flaws. If you use this metric, there are chances that you will pick a winner who doesn’t deserve to be and this will be an injustice to the actual winner who should be awarded the prize. There are a couple of reasons for this.
Why is pageview not a very good metric for picking the winner?
The biggest problem with pageviews is that they will never tell you if the article was helpful to the reader or not. It will only tell you how many people saw the article. An author can have the most views for an article but the article can be mediocre and can still win the prize when they shouldn’t. The reason is when we see a link we don’t know anything about what it contains, we decide whether or not we should read the article based on only the headline of the article. And as soon as someone clicks on the link and visits the website, the pageview for that article gets incremented by one. So, we are actually giving credit to the author before the reader even reads the article. Which is not a very good way to measure the quality of an article? We can do much better than this.
The best metric to measure the quality of an article is the page value. Page value is the average value for a page that a user visited before landing on the goal page or completing an Ecommerce transaction (or both). It measures the influence of a page and helps us better understand which pages on our site drive more conversions and revenue. So, if an article generated $10 for Dataquest, another article $5, and one $0. Then it’s not very hard to choose who is the winner. There is no ambiguity in choosing the winner and nobody can question your method because at the end of the day the goal is to make more money and reward the articles or any activity that helps us do that. But if you use pageviews, then there will always be some ambiguity and anyone can question your judgment because no matter how many times you beat pageviews with a stick, it will never tell you if the article was useful to the user. You can think of pageview as a proxy for measuring the attractiveness of the headline of the article but not the article itself. So, I think using the page value would be a better choice to pick the winner and ensure that the right person gets awarded for their hard work. Recently I built an ml model to predict which customers will buy from an online website based on their activities on the site. I found that the most important feature for predicting the outcome was in fact page value. If you want to look at the model, you can do it here - https://customer-purchase-prediction.herokuapp.com . Note the dataset doesn’t contain pageviews so I can’t compare between the two but you will get the idea of why page value is such an important metric.
I hope this information is helpful to the community. Thanks for listening.