Skip to main content

A touch of magic to transform retail operations

(Image credit: Image source: Shutterstock/Maxx-Studio)

When stores where you regularly shop, run out of stock on your favourite or routine items - is this a source of minor inconvenience - or an unthinkable outrage?

In today's digital age of instant gratification and plethora of channels to shop consumers are more likely to act with bemusement bordering on horror when a retailer cannot immediately supply the items they want. Consumers don’t have the need to sympathise or it’s not the consumer’s fault for being beaten by the rush, but rather the retailer’s for failing to predict and plan for the unexpected demand.

And that’s exactly as it should be. Every retailer should aim to have the right products available at all times. With so much competition in the retail market, businesses that fail to give customers what they want will quickly lose both reputation and market share. This means retailers should have a clear grip and visibility on inventory, globally. This means that the operational efficiencies must be enhanced for running an effective supply chain. Right from the ability to plan ahead for a peak season or an event to tweaking the stock based on real-time demands is becoming a norm. It is imperative to run a tight ship and avoid excess or less stocking. Global inventory visibility has been an issue for many Retail & CPG clients. Clients have tried to tackle this problem by bringing in operational processes to consolidate the channel and market view. However, given that businesses are now hyper-agile the operational processes are lagging and there is a need to optimise the process and also have a technology solution that provides for real-time or short latencies.

A complex machine 

With the increasing complexity of retail operations, every cog in the machine, from manufacturing to supply chain, stock management to dynamic pricing, depends on the ability to forecast demand accurately.

And while it’s relatively easy to pore over historic data and identify the reasons why certain products were so popular at various times (for example, because of a heat wave or a new sales promotion), this insight is not enough on its own. Retailers must move incredibly quickly to implement these findings across every area of their operations, from the warehouse to the point of sale, if they are to react effectively to changes in demand.

This requires retail businesses to crunch enormous volumes of historic and up-to-the-minute data, while also automating processes throughout its operations to ensure that they can quickly implement the right changes to fulfil orders and keep the shelves stocked. A technology platform that provides for quicker data ingestion, effective storage and retrieval of insights from this vast volumes and variety of data is the need of the hour. Traditional ways of running business will not cater to the changing times which requires processing at the sources or end points.

Software and silos

Retailers need to revolutionise and replace many long-standing approaches to business information. The biggest barrier to implementing an effective demand forecasting system is the traditional silo-based approach to data, which leaves this valuable information trapped in a multitude of separate applications.

The goal for retailers must be to break down the walls of these silos so information is shared quickly and efficiently across the business. This depends on effectively managing the many different sources from which data is gathered, transforming data processing and storage so that it is available to every line-of-business application that needs it, and introducing agile, need-based processing that requires no intervention from IT. In simple terms, an effective data governance strategy coupled with the technology platform is critical to run business effectively.

Modelling the future

Once the retailer has a handle on their data, they can then begin analysing the many variables that affect demand. This requires mapping the relationship between sales and factors including (but are by no means limited to) seasons, holidays, promotions, times of day, and previous customer behaviour. Using exploratory data analysis (EDA), the business can break down massive data sets, construct and test a number of predictive models.

Through regular analysis, retailers will soon establish patterns and relationships that will enable them to identify and forecast trends that have an impact on stock-keeping unit (SKU) availability, and then build algorithms that automatically make the necessary changes within the relevant areas of operations, from supply chain to pricing. The technology platform alluded to earlier should cater to the twin forces of provisioning & consumption. For any such analytics to be performed, more than half the time today is spent on data preparation than on actual algorithm, training or insights generation. The platform should provide for effective data wrangling and modelling abilities.

Any sufficiently advanced technology is indistinguishable from magic, and there is something magical about predicting the future of retail demand. At Infosys, we have seen retailers gain the ability to predict sales by channel, product category and time period with more than 85 per cent accuracy. This insight gives them significantly better visibility in short-term order volumes which enables them to streamline budgeting, sales planning, warehousing and fulfilment.

Not only does accurate demand forecasting translate into more efficient operations and improved revenue; it also ensures that retailers should never have to face the embarrassing, reputation-damaging impact of empty shelves and frustrated customers. Infosys offers a wide range of solutions that takes clients through the transformation journey rapidly. Read on the solution offering details on ‘Information Grid’, ‘Analytics Workbench’ and related to get a perspective on what is in it for me and how do I get started. 

Modernising to aid monetisation is imperative for Retail & CPG clients to thrive in this dynamic market state. So, why wait…

Ramakrishnan Krishnan, Associate Vice President and Head of DNA Retail, CPG and Logistics Delivery, Infosys
Image source: Shutterstock/Maxx-Studio

Ramakrishnan Krishnan is a seasoned veteran in the Data Analytics space. He has consulted and delivered many DWBI engagements for various clients across the globe. Currently as Associate Vice President, Krishnan is the DNA Practice Head for Retail, CPG & Logistics industry segment and also heads the Data Management practice for Infosys.