Productivity Hack Thursday: Optimizing your Data Analysis Process

If you are a data analyst (we all are to some degree), your work typically comprises 4 steps:
– Data collection
– Data cleaning
– Data analysis
– Data interpretation

Assess the time you spend on each of the four steps of this task. Where are you spending most of your time? If, for instance, you had 10 hours to complete a data analysis exercise, how would you split your time across the four steps?

The split of time spent across the 4 steps can signal the level of maturity of your data analysis process and the opportunities for improvement from a time usage perspective.

If you find yourself spending more time on the first 2 tasks than on the latter 2, there are strong opportunities for optimization. Data collection is a low-value task that can be easily automated for routine reports, so time spent on this step should be as minimal as possible. Data cleaning is a more specialized task, but spending a lot of time on it means the source data is not being entered or retrieved in an ideal manner, and there are opportunities to optimize the data entry or sourcing process. Data validation, controlled data access, and data filtering are a few of the ways the data sourcing/entry process can be improved to reduce time spent on data cleaning.

There is a need to identify the most appropriate analytical tools or methods or charts for the task at hand, so it is understandable if time is spent on the Data analysis step. However, this time is expected to reduce for routine reports, as the decision on analytical methods would be fixed.

As high-value knowledge workers, the bulk of the time is expected to be spent on the final step – The interpretation of Data. Data holds a lot of treasures – useful insights for business improvement and growth – that need to be mined out through comparative and correlative assessments and other cerebral activities that will require dedicated use of information processing resources.

Certain instances might require scenario A or B for best results. However, for routine data analysis work, scenarios C and D are most ideal. More time spent on analysis and interpretation guarantees more useful and actionable insights derived from data, which leads to better business decision-making. And this is the desired outcome of the data analysis process.