Optimal version release strategy for mobile apps

Abstract

There has been a recent rise in the number of mobile application (popularly known as app) based startups. Also, ventures which earlier relied on website are fast switching to mobile app only model. Uber Cab, Chai Point are only app based while others like Flipkart, Amazon and other e-tailers are offering discounts to users for using app instead of the regular phone call or website purchase. Other than the reduced overheads of call centre a major benefit in mobile app based transactions is the possibility of targeted advertising. Thus in-app advertisements have also become a major source of revenue for some apps due to the availability of user information such as location, purchase history etc. This high utility of apps has led to competition among companies to provide the most userfriendly features and attract customers. For example, Taxi for Sure, an aggregator of taxi service in Indian cities has started offering magic rides feature in its app wherein the app suggests the most used trip. Features such as these involve cost and also time of the product development team and decision to implement it needs to be checked against the possible increase in customers and revenue. The decision to include any feature in a release of the product is crucial as considerable investments have to be made in developing it and the costs recovered from the additional sales. The product manager may opt to launch a very basic product at first and based on its response decide to include additional features in subsequent launches one-by-one or in some combination. This project aims to develop a tool to aid this decision making process for the product manager. Due to the uncertainty in market response a fixed strategy of version releases is not a dependable solution. The tool should provide capability to the product manager to change the launch strategy after release of initial version based on its market response and select the optimal version(s) to launch at that point. Dynamic programming provides a suitable platform to address such changing scenarios and therefore the dynamic decision making tool is developed using dynamic programming.