Business intelligence and analytics of the Italian automotive market : big data applications in GoodBuyAuto

This document presents the analysis of the Italian automotive market and car pricing methodologies based on the use of Big Data analytics. This work was realized in the company GoodBuyAuto, an Italian startup in the online automotive marketplace, as part of an internship done with the Politecnico di Milano. This project is based on the premise that Business Intelligence, by the use of huge amounts of data (Big Data), focuses on analyzing it for obtaining valuable and relevant information that can be used for making more efficient and effective business decisions, thus driving competitive advantages into businesses. The most important question to be answered by this work is how to correctly price a car?, based on different characteristics it can have and different aspects in the market that can influence its price setting, as viewed from the point of view of the sellers (asking prices). Different tools of extraction and data analytics have been used and analyzed through the development of this project. The final solution is presented as the use of Tableau® (a powerful software used for data exploration and visualization) and R (an open-source programming language and software environment used for statistical computing). In this way, different solutions regarding Business Intelligence and Analytics algorithms have been deployed for GoodBuyAuto, which directly impact its day-to-day business operation. The most important solution implemented for GoodBuyAuto relates with a direct application of a supervised machine learning algorithm: a multi-variable regression predictive model based on different car's characteristics, in order to make a detailed price estimation and evaluate in this way possible business opportunities for profit maximization (market abnormalities in the sense of low car pricing). This solution is composed by a set of three main interactive dashboards that allow the users to make, in a general and sequential way, a detailed data exploration and statistical training/evaluation of different predictive models in order to realize a final car pricing according with the filters, settings and input variables previously selected. The second solution is related with different perspectives for market analysis (that can be reduced in the end to the combination of three main dimensions: time, price and location), and the creation of actionable and ease-to-use dashboards that convey direct business information and valuable insights for for GoodBuyAuto's management and its main business departments. The final point is to highlight the importance that this kind of solution can have for the company. Decision support systems based on data analytics can provide important insights, and decision driven information that can lead to competitive advantage and to the creation of business value from the appropriation of the insights/information obtained from data analysis. The importance of this kind of solution is clear and the need for actively investing becomes more evident as data and technology become more accessible. Some suggestions and future work activities are proposed for GoodBuyAuto so their Big Data infrastructure can be further improved and much more hidden opportunities within data can be discovered in the near future.