Exploratory data analysis (EDA) is an important practice for getting familiar with a dataset, but it can be hard to know at what point an exploratory analysis is "done". Striking the balance between diving deep enough into a dataset so that you can be confident about an analysis, but not diving so deep that you're spending extra time without additional value, can be tricky.
By predefining success, we can optimize our exploratory data analyses to balance value with time spent. This talk will focus on strategies for optimizing the EDA process to make analysis as impactful as possible with a focus on how and why to define success up-front. Once we've defined what success can look like for EDA, we'll walk through how to use success criteria to design an effective analysis that deliver value in addition to helping us to better understand our data.