Retail

How can retailers incorporate consumer insights to enable merchandise optimization?

scroll down

Amitabh Mudaliar

Business View

Amitabh Mudaliar, Vice President & Senior Client Partner, Retail, CPG & Logistics, Infosys

"Many large retailers have an internal consumer insights team in place, but they often work in silos. As a result, insights do not end up being actionable. "


While planning merchandise for brick-and-mortar stores, in the past retailers relied heavily on data from multiple sources, including historical sales data and other market data from sources like Nielsen. Though there were tools to analyze data, most of them did not leverage very advanced modeling techniques. Today, however, retailers operate in a very different environment – one that is driven by two factors. The first is the emergence of multichannel-enabled consumers that presents unique challenges in merchandising across channels to a highly informed consumer. The second is the widespread availability of tools that leverage advanced modeling techniques to analyze data for consumer insights and planning. We see retailers still figuring out how to operate in this environment.

To thrive, retailers have to take a three-step approach. First, they need to ensure that they have the right tools in place that leverages the science of today – tools that leverage advanced machine learning algorithms to drive planning and insights. Second, they must re-look at their merchandising processes – are consumer insights being effectively leveraged in the process for merchandising across channels? Thirdly, it is important to ensure that the organization is aligned - many large retailers have an internal consumer insights team in place, but they often work in silos. As a result, insights do not end up being actionable. Often a reorganization is needed to ensure that the two teams (consumer insights and assortment/merchandising) benefit from each other.
Alex Farcasiu

Technical View

Alex Farcasiu, Senior Principal – Analytics and Information Management, Retail, CPG & Logistics, Infosys

"Machine learning and artificial intelligence have made significant strides in last few years, reaching an inflexion point."


There are three major technology shifts that are contributing to the realization of consumer centric merchandising vision. The first is affordable analytics. analytics technology (both hardware and software) are getting cheaper year after year while also improving in capabilities; making Big Data analytics mainstream and bringing retailers closer to the point of truly understanding the customers and their preferences and factoring those inputs in key merchandising decisions.
Then there's the hybrid database – an advanced merchandise planning solution requires both fast reads and fast writes at the same time. Traditionally database technology used to be good at either fast writes (OLTP) or fast reads (OLAP).The rise of new generation database technology called HTAP (hybrid transaction/analytical processing) like LogicBlox that offers both fast read and write capabilities is changing the game. Finally, machine learning and artificial intelligence have made significant strides in last few years, reaching an inflexion point. Both unsupervised and supervised learning are proven to be highly effective in solving the hard problems in merchandise planning, like predicting which new fashion items are likely to be hot sellers this spring. Together these technology shifts are demystifying the 'art' side of merchandising and pushing the boundaries of what is possible.
Shachin Prabhat

Domain View

Shachin Prabhat, Senior Principal – Supply Chain and Merchandising, Retail, CPG & Logistics

"Analytics, especially consumer analytics, should be the fulcrum on which all merchandising functions should rest, be it store clustering, assortment, pricing or promotion."


Today, retail has fundamentally changed. Rise of digital consumer and growth of omni-channel commerce is pushing retailers to rethink their current merchandising processes. The key to succeeding in the new world of where the consumer is king is to transform technology, processes and organization.

Specifically, retailers of today should embrace the new technology. Maturity and affordability of modern technology constructs like Big Data analytics, machine learnings etc. are making it possible to bring a rigorous data driven approach to merchandising. Rather than rejecting the new technologies, retailers should embrace and leverage them to revitalize the merchandising landscape. Next, they must make analytics the core of merchandising. Analytics, especially consumer analytics, should be the fulcrum on which all merchandising functions should rest, be it store clustering, assortment, pricing or promotion. Today, at most organizations, consumer insights is a treated as a side function operating in silos. Instead, it should be deeply integrated in merchandising landscape. Finally, retailers must zero base the merchandising organization. Current merchandising organization at most retailers is based on the business model of the last century. There is a clear need to relook at the merchandising organization and align the structure and incentives to match the new realities of cross-channel retailing.