1.) Obtain data in form of structured, un-structured, semi-structured etc. In any type, SQL, NoSQL, .CSV, Excel, JSON etc.
2.) Clean the data either ETL or ELT (Extract Transform Load or Extract than Load and than Transform)
3.) View the data, check its "value, variety, veracity, volume, velocity".
4.) Model the data as per requirement, so that it can be process as per data model.
5.) Interpret as per analytics or as per analysis of that modeled data.