Tuesday, April 14, 2015

Context + Quantitative Data = Correct Prediction




Now, we have access to future trends based on historical and real-time data. We can now predict at least something about our customers’ future actions. However, we are not sure that our predictions are 100% sure; there is always a clause. Why? 

It is because that quantitative data alone cannot give the 100% answer to our future queries. We need context related data which we most often won’t get. For example, if a customer of a retail chain suddenly triples his purchases for 2 months; then what do you predict based on the data that you have? Is he going to buy at the range in the third consecutive month also? Or is it just a fluke purchase spike which you can ignore? Or is it something like that the customer’s family has been extended recently and due to this his spending pattern has changed drastically? Which one of the above questions correctly predicts the customer’s behavior? 

Quantitative data only provide the future trend based on the purchases that the customer has made so far; it won’t collect the reason behind this sudden spike in purchases. In this scenario, to get a comprehensive understanding, you need to have context related data along with the regular quantitative data; only then your predictions about the future purchases of the customer holds true up to a majority of the extent.

What do you say?     

No comments:

Post a Comment