Visualize the Correlation of Variety of Big Data

By Suresh Acharya, VP Product Development, JDA Software

Suresh Acharya, VP Product Development, JDA Software

Big data analytics can be applied in supply chain to derive tangible benefits but sometimes what becomes prohibitive is our perception of data. Data is not quite what most of us think it is. Data is the fuel to the models and whether or not it will show something which we want to see is the issue. Big data is bound to become more and more important and the companies who have successfully used big data continue to be successful.

When we talk about big data, there are four ‘V’s namely Volume, Velocity, Variety and Veracity. Volume is the petabytes of data that is sitting there for us to mine. Velocity is the rate at which data is coming in and Variety is the different types of data and how to get information from it. Lastly, Veracity is related to the situation when you have incomplete data.

The significance of these aspects can be understood through an example. Let us say

you are running a promotion of 20 percent off on clothing. You receive a stream of Point of Sale (POS) data and realize that your promotion is doing better than what you predicted. But, what if you are running the promotion for three days and you don’t have enough inventory. It is quite likely that it is going to run out on the very first day, and you do not want your customers to come and see an empty shelf.

So, what you see there is the volume and velocity of data coming in. You need to be able to mine that data for real time prediction. However, we have considered only volume and velocity of data and not the variety. Have you looked at the weather data? The weather for Saturday might be terrible which may be the reason why everyone is buying on Friday. So, the fact that you get good velocity on Friday is not a representation of the ongoing promotion on Saturday and Sunday. This is where the variety comes in. Your POS data hits an orthogonal weather data which is independent. Another orthogonal dimension can be the social media. So, now you have different streams of data and you can mine them independently and correlate the data to come up with reliable predictions. JDA’s role in this market is to visualize this correlation.

If your supply chain is nimble enough, you can also take advantage of real time analytics to place an order by detecting a potential stock out. You can get the promotional stuff sooner and the optional ones can come later. This is where real time analytics of seeing the POS and understanding how that correlates with the other insights enables you to do a better job on the supply chain side.