B&E | Predictive Analytics in "Unified Commerce" - A Brief Insight
IndraStra Open Journal Systems
IndraStra Global

B&E | Predictive Analytics in "Unified Commerce" - A Brief Insight

By Rahul Guhathakurta

Globally, "Retailing" has been always on a high-octane mode from the inception of Walmart in the 1960s to the launch Amazon in 90s. But, as we ushered into the digital landscape with new knowledge on customer data, pertaining to its behavior  - the timeline of decision makings and executions has been drastically shortened. The arrival of "Unified Commerce", armed with "Predictive Analytics" support system is acting as a catalyst for the further growth of retail in multiple channels and formats. But, how it is going to work?

"Unified Commerce"? What? 

Till 2014, everyone was talking about omnichannel retail at any given NRF (National Retail Federation) events. In omnichannel, one has multiple channels and multiple formats, but don’t have one transplatform software and supporting network. But, in the "Unified Commerce",  all the channels of retail operations are connected in real time - both physical and digital, constantly churning out raw data for the authoritative analysis purpose.    

When it comes to generating customer intelligence in "Unified Commerce" business model, a structurally stable triangle methodology has to be deployed. Each of the three angles is critical to the triangle for retaining its structure and fulfilling its intended purpose. The three "angles" of the triangle are "retail domain experts", "technology experts", and "data scientists".

Image Attribute: Channel Evolution - Unified Commerce / (c) 2015 Boston Retail Partners / Source: Slideshare

Image Attribute: Channel Evolution - Unified Commerce 
 (c) 2015 Boston Retail Partners / Source: Slideshare

What is  the fuzz about "Predictive Analytics"? 

"Predictive Analytics", per se tends to bring in crystal clear clarity with respect to KPIs in retail and with time, the overall system will evolve  with the advancements in technologies like in the field of “Internet of Things”. In real-time, targets can be set at different nodes of supply chain channels  through the insights generated via analytics. These targets may form a part of the management-by-objectives in an organization, along with pre-defined  tolerance factors. Now, when an actual value exceeds one of these tolerances, a corrective trigger will get fired seamlessly across the platform.  

However, the "accuracy" of such predictive insights will always be a hard hitting question and it completely depends on what kind data one is sourcing and harvesting through these inter-connected systems at the first place itself.With time, the quantity of data will grow, so will be the discrepancies. In such scenario, a well integrated  predictive analytics module will act as a system of "checks and balances", providing real-time impetus to managers to deploy corrective actions wherever and whenever it's required.  

Predictive data scientists are then needed to use the stored data to build models that achieve those business objectives originally set by the retail expert derived from in-store planogram execution, average sales value, gross merchandize value, average order value, inventory and supply chain costs, etc. Predictive models find relationships between historic data, subsequent outcomes,  consequent measures, and counter-measures - so that near-term and long-term customer behavior can be predicted. This "angle" of a triangle is an answer to problems such as the probability of when a shopper will make their next purchase and what the value of that purchase will be. Gradually, these relationships will tend to become more complex problems that only machine learning techniques will be able to solve. 

The "Call for Collaboration" and Why? 

But how do "data scientists" determine which derived data outputs are relevant? Usually, data scientists lack the deep domain expertise needed to clarify and prioritize their interpretations. Therefore, a collaboration with "retail domain experts" is essential.  And, to make that data available from customer's end through well-optimized systems, calls for the presence of "technology experts" in the equation of collaboration.

What Future Holds?  

It's indeed a high time for retailers to rethink their technology approach with regard to an ever-evolving crop of "connected customers" who are already accustomed to "unified banking and payment system" since 25 years. The bottom line is, wherever money gets transacted, a  retailer should have its POS deployed at that place, it can be either physical or digital.  Yes, there will be the challenges to fight, acquire and retain the customer's short-span attention. But, according to the Pareto principle of 80:20 - the 20% of customers always provide the 80% of sales. That 20% is changing constantly, and to be able to identify that 20% of the top line customers in real time—and manage through extraordinary line of services—is really the "holy grail" of successful & sustainable "unified commerce".  

About the Author: 

Rahul Guhathakurta (TR RID: K-4094-2015) is the founder of IndraStra Global, a strategic analysis, and information services provider. He distributes his time between New York and Ahmedabad.