Data analysis needs to be fun

2016-03-21 1 min read

    In order to do any meaningful data analysis you need to have fun doing it. Otherwise it becomes a chore that’s extended by each additional analysis you run and each additional failed attempt at an insight. This requires a positive attitude and enjoying the slow, methodical process of discovery and appreciating each iteration while getting closer to the end goal. The vast majority of analyses lead to no new insight, especially when all the easy stuff has already been figured out, and it’s critical to remain the optimist while appreciating the present.

    A key part of this is tools. I have a set of tools I’m intimately familiar with and can manipulate them without much thought. It’s this passive approach and behavior that lets me go through the rote work while simultaneously focusing in on the challenging elements of the problems I’m facing. Fast tools are also a requirement. Tools that allow you to quickly get a result prevent you from leaving the zone and leave you ready for the next attempt.

    When 80% of the work is rote data manipulation it’s important to not burn out while getting to the 20%. To be successful you need to find the fun in both the data manipulation and the analysis.