ANALYTICS

SECTION ONE

SECTION TWO

Assessing the Opportunities of Big Data

By Mark Osborn

Big Data refers to the ever increasing volume, velocity, granularity and variability of data now available. For consumer products (CP) specifically, Big Data represents an enormous opportunity to leverage new and different types of data to accelerate consumer-driven innovation, rationalize sourcing and procurement, optimize manufacturing, streamline supply chain and reach, and engage and serve customers and consumers alike. To begin to assess the opportunities associated with Big Data, let’s consider each of its dimensions.

First, the volume of available data is increasing exponentially. Not long ago consumer products companies made do with a specific set of data: internal financial data, order and shipment data to drive periodic demand and supply forecasting, and capacity planning and inventory management data. Today, in addition to those core data types, consumer products companies have new-found access to retailer point-of-sale (POS) data, retail measurement syndicated data, supplier inventory data, real-time manufacturing capacity and inventory data, social media monitoring, periodic consumer research, and more.

What’s more, this volume will only continue to increase. The coming age of the Internet of Things and eventually the Internet of Everything, will create even greater volumes of real-time streaming data interactions between sensor-enabled devices, machines, locations (retail stores, for example) and consumers. The incremental value of this data will be enormous, representing significant new opportunities for CP companies to understand consumer behavior, anticipate consumer needs, and engage with consumers in real-time. However, the corresponding data management requirements will be substantially more significant than even those requirements that are emerging today.

Next, data availability is accelerating to approximate real-time across a wide variety of data types. The ability to acquire, harmonize and integrate increasingly large volumes of data at faster and faster cycles is becoming critical to enable fast response to consumer needs, preferences and market demand dynamics.

The velocity of data is also serving to accelerate other business processes such as planning.  Companies are increasingly leveraging the faster pace of data to accelerate planning cycles, to the degree that what was once point-in-time forecasting is now evolving to real-time, rolling short-term demand anticipation.

In addition to volume and velocity, data is increasingly available at lower and lower levels of granularity, creating incredible opportunities for finely-grained analysis and planning against, for example, increasingly refined segments of consumers, region-specific seasonal SKUs, and more.  The ability to plan and execute against increasingly granular data sets improves consumer products companies’ abilities to target consumers, tailor products and offers to local markets, develop retailer-specific programs and promotions, and monitor other key business processes for risk mitigation.

The variability is also challenging on a number of dimensions. First, the timing of data acquisition and delivery can vary dramatically by data type. For example, some data elements might be available in real-time, while others are daily, weekly, monthly, or ad hoc. For a company trying to develop an accurate view of current demand, assessing actual consumer sell- through when, say, 10% of the data is real-time, 40% of the data is available daily and the remaining 50% is available a week, a month or more after that demand occurs, bringing these varying data elements together in a meaningful way and accounting for the differences in data availability is a significant challenge.

 

Next, associating and correlating varying data levels can also prove difficult, independent of the availability of the data. For example, sales managers forecast in terms of cases, while marketers in terms of SKUs, demand planners in terms of units, and financial managers in terms of revenue. All are indicators of demand, but bringing cases, units, SKUs and revenue together into a consensus view of demand based on a commonly understood number that can be related back to each group’s “own” definition of demand can be daunting.

 

Finally, companies are further attempting to correlate quantitative transaction data with qualitative data such as social media monitoring and consumer research. However, correlating changes in consumer perception with a spike in order volume or a price change at the retail shelf, for example, requires significant data management and advanced analytics capabilities.

Mark Osborn is VP, Digital Strategy and Business Planning at SAP Consumer Products Industry Business Solutions. For more information: www.sap.com

Click on the LinkedIn logo to join the new Shopper Technology Institute Discussion Group

SECTION THREE