Leveraging Social / Consumer Genome for Merchandising
Most retailers traditionally leverage sales data from POS terminals to analyze buying behavior. In some cases, loyalty card data is also used to determine appropriate assortment decisions. These data sources and their corresponding analysis have proven reasonably helpful, though they don't convey the whole picture. With the recent explosion in social and consumer related data on the web, there is a wealth of information that Retailers should be exploiting and incorporating into their merchandising decisions, primarily around assortment and space.
Social/Consumer genome related data provides rich insights into needs, wants and buying behavior of individuals. This data when combined with demographic and geographic data can provide a good map of consumer wants and needs for a given market for a set of product categories.
A major input into merchandising decisions / assortment plans is to determine consumer buying behavior to identify what products sell and what potential products could sell to increase sales. Based on this analysis, assortment decisions of inclusion/exclusion or allotment of space are provided. The means to identify this buying behavior was typically the use of POS data or data from Nielsen/IRI that provided good basis for WHAT was being purchased. When married with Demographic data, there was a good proxy for WHY the products were being purchased. Even when rigorous correlation and clustering analysis is carried out, the determination of buying behavior and the reasons for the same were proxies at best.
Now with the availability of consumer genome or social genome information, the analysis of WHY purchases are being made and what is being purchased with identification of latent and express needs becomes even more accurate as there is clear expression of wants and needs. This will significantly enhance the quality of assortment and merchandising decisions as the degree of error/approximation is reduced.
There are however some pitfalls to the use of this data. We cannot solely rely on this data as web usage and consumer/social genome information may not fully represent buying needs and wants for the entire market population. This data usage has to be married to traditional sales analysis to augment the decision making process.
There are no tools / application products in the market place that provide truly integrated capabilities. The holy grail for optimized merchandising would be to integrate social/consumer genome data effectively into the traditional clustering analysis and thereby into the assortment planning process. This might be a challenge for some of the retailers who struggle with traditional approaches. Asking them to adopt more advanced analytical approaches to incorporate social/consumer genome data would be a challenge.
The key would be to devise suitable technology/process platform augmented with robust analytics shared services that can leverage necessary data to enable optimized merchandising decisions.
This article has been contributed by Amitabh Mudaliar (Group Engagement Manager - RCL Infosys). You can reach Amitabh at Amitabh_M@infosys.com.