Is Clustering another name for Segmenting?
Talking about the famous automobile. - Model T - Henry Ford once remarked, “Any customer can have a car painted any colour that he wants so long as it is black.” Those were the good old days when a marketer could offer a single product and sell it to everyone in the market.
Cut to our times - we have products in every conceivable shape, size, colour, flavour, and every other possible variation that one can imagine. As a customer, you can be as specific as you want to be, and there’s a good chance that there’s a variant to cater to your tastes. The market, therefore, is divided into many smaller “markets” that have identical tastes within themselves. These smaller “markets” are more accurately called “market segments”.
Marketers love segments; these give them a chance to capture a base within the market, all to themselves. Segments allow a marketer to create an offering just for the tastes of a segment, distinct from the offerings for the other segments.
In predictive analytics and exploratory data mining, we often hear about clustering. To put it plainly, clustering is a process of dividing all objects in a set into a certain number of groups. So, what’s exploratory about it? Before performing clustering, the analyst isn’t aware of the total number of clusters that would eventually result. This creates room for exploration. One can cluster the same data in different ways, into different numbers of clusters, and garner different insights each time.
After the prospective customers in a market have been grouped into clusters, the marketer may be able to find some commonalities within each cluster. Knowing the product along with the groups in the market, it’s possible to identify the groups that would be most interested in the product. Similarly, it is also possible to identify the groups that would not be served by the current product, but may be served with a modified or even a new product. That is the start of segmentation, because the marketer can then offer differentiated product offerings to each segment. Of course, it’s also entirely possible that a marketer decides to not participate at all in a particular segment.
Clustering is one of the techniques to get started with segmentation. Clustering algorithms are especially helpful when trying to analyse huge amounts of data and group them into clusters. The difference between the two lies in their utility - clustering helps to identify the groups, while segmentation is using the inherent differences to serve the market better.
>>To know all about Business Analytics with R, please go to Manipal ProLearn.
Soumyadip Pal is a retail analytics professional and a passionate educator with more than 8 years in the industry and more than 7 years in the academia, currently working as a consultant with Manipal Prolearn.