The data scientist needs to be versatile. On the contrary to the data analyst, the data scientist examine not only the data of a single source (ex: CRM) but the data of numerous sources. He has to extract the main data and define the concrete indicators. He analyzes both internal as external data (customers, prospects, employees) that the company gets by various ways.
Then, he makes recommendations to improve the product or the service of the company, or its performance. The missions of data scientist are various. He has to begin with a "business" issue of the company and to translate it under a statistical shape. He has then to find the sources of data to answer the problem.
Finally the last mission of the data scientist consists of the analysis of data and the emission of strategic recommendations on the database (modify, repatriate, to outsource or to internalize).