Data Insights: Overview

In the last two segments, you learnt about two methods through which you can generate new information from a given dataset on which you can perform the five patterns to gather insights. It may not be evident right now, but these two analysis techniques are quite powerful in generating insights when you are stuck at a certain point and feel like you have done all the possible analyses. Let’s hear about Anand's views on this.

Now you may have a question in your mind - “I’ll be able to generate hundreds of insights from the data by repeatedly following the procedure given above. But how do I know which ones are useful for me? That is how do you know which insights should be focussed more than the other ones? Let’s hear from Anand on how he differentiates the insightful results from the non-insightful ones.

So now, you have a particular checklist through which you can go through the entire dataset, do a variety of analyses -either by collapsing the rows or preparing new columns and from the results infer whether the generated insight is useful or not.
The qualifying criteria for insight are given as follows:

• Is it useful?
• Is it interesting? 

Also, as mentioned in the video, it is always a good idea to have a list of hypotheses ready with you when you’re preparing to analyse certain data. You can create them on the basis of the five patterns that you learned previously or you can utilise the business aspect of the problem at hand. Combining both of them, you generate insights and see if it leads to something useful and interesting.