A successful data scientist should have the ability to translate business objectives and problems into analytics problems. If we have unlimited sources of data, an interesting question arises that what data should be chosen to predict which business or analytics target. We can always get data science project ideas from tech companies’ blog and Kaggle. Even though we don’t have the data and resources to work on it, it’s still important to get the idea and hopefully transform it into other practical situations. This post will show you different data choice to tackle different business concepts.
Customer Lifetime Value( LTV)
LTV is an important factor used for making decisions like promotions, subscription fee and so on. Airbnb used crowdsourcing to create features at different granularity such as hosts, listings. Another project from Kaggle that a payment company called Elo, used merchants, card info, historical and new transaction data to predict LTV. What’s more, from an interview I personally experienced, they assume the lifetime value of a customer is equivalent to the purchases a customer makes over a given timeline and group the customers into cohorts and keep track of them. Their first-time transaction tells you which cohort they belong to.
I will continue to update this post.
