As the typical scenario in any data analysis includes more than one type of data source, working with large datasets, messy and unorganized data, there is a huge need of data prep required. Most data sets are relatively dirty and need to be thoroughly cleaned for the analytic result to be usable. The need to have some structure for reporting and analytical tools to grab onto resulted in a boom of data prep.

It is very imp to have the data validated in the initial stage, because if that goes wrong, then everything downstream of that becomes very problematic. Thus we need to have the data ready for analysis and to avoid any non-value add, which is achievable by big data prep.

Big data prep uses a combination of machine learning algorithm to automate most of the work that goes in sanitizing data. Read more here- http://www.datanami.com/2015/06/22/why-big-data-prep-is-booming/