With the evolution of technology and growth of data, The term Analytics has derived several meanings over the years. Depending on the type of data, size of data, method of working with data, type of queries that needs to be answered, Analytics has been grown into several types.
The magic about this field is that the resources available are unlimited. Be it the data that we need to work with or the tools we need to use, We have TONS of ways of going about it. It’s a blessing and a curse. It’s a blessing because we have not one, but MANY ways of working with data. The magic is, THAT’s the curse too…because we then have one too many ways of getting confused. That’s why, to avoid all this confusion, the field of analytics has been generalized into four different types based on the questions we need to answer.
The first thing you need to do when you start working with data is to scan the data into your mind, understand the data in depth. Doing that is of course, practically impossible, especially when you’re working with large amounts of data. We need to breakdown the data into capsule version and study the data to understand what’s been going on and to derive what needs to be done. Doing this is called descriptive analysis.
When you inform your doctor about your persistent problem, he or she is going to want to do some tests to figure out what’s causing the issue and why.
When you start doing that with data for an organization, that’s when it’s called diagnostic analysis. Here we look at all the data we have had in the past and all the data we have now, conduct a detailed examination to figure out why you have got the outcome you have now.
How is it different from the descriptive analysis? Well to put it simply, Descriptive is when you work with the data at hand and diagnostic is when you work with all the past data and the present data to arrive at a solution.
Naturally, this is the next big step after diagnostic/descriptive analysis.
Once we study the data, we can use certain steps (complicated statistics comes into play here) to analyze/predict what might be the outcome for the coming years.
For example, If a clothing retailer stores up and analyzes all the sales data of the past few years, he can figure out which material, which brand, which type of cloth was sold at what rate during what season. Like thicker woolen clothes would be sold maximum during the winter season and cotton types during the summers.
This helps us learn the different factors that might affect our outcome and what needs to be done to increase our efficiency. Of course, the best accuracy is obtained when high-quality data is used, and also, this is the phase where machine learning, deep learning, etc come into play.
Once we have studied the data, predicted several outcomes with our beautifully built models.. what’s the next step? YES, it is to find the best solution to our problems. The prescriptive analysis is exactly what its name suggests. It prescribes us possible solutions that need to be strategized and implemented to help with the growth and development of the organization.