Making Sense of Consumer Data in the World of Retail

 Last week we looked at an example of the consumerization of big data from the publishing world, examining the relationship between brands, consumers, and their data. Today we take a look at the retail industry and the ways retailers can use data sets to anticipate customer behavior.

Shish Shridhar, Retail Industry Solutions director at Microsoft, demonstrates how retailers can predict buying habits through a combination of their customer purchasing data and various APIs that provide additional, correlative data sets. Take, for example, an analysis Shridhar includes in his presentation of two grocery store chains in the city of Seattle: QFC and Safeway. Using Bing Maps, Microsoft Excel, and Seattle city data (, Shridhar shows us how sales data can be correlated with demographic data for each store location.

This ability offers individual retail locations a key competitive advantage because smarter data means increasing profits by accurately predicting behavior. This allows stores to make data-based decisions around which products to stock, which in-store deals to offer when, and which coupons to mail to their customers. In short, they can leverage the data into business insights that can directly affect their bottom line.

How to ensure an accurate read of your data to create actionable insights

These are the sorts of business insights that big data can provide. However, while we have access to larger and more diverse data sets than ever before, identifying customer insights and showing a causal relationship still requires human input and manual analysis. Shridhar warns against one common statistical pitfall when it comes to data analysis: “correlation is not causation.” This maxim is familiar to scientists, but may not always be familiar to business decision makers.

In his presentation, Shridhar suggests three methods for ensuring an accurate read of your data to create more actionable insights:

  1. Avoid bias. It’s easy to begin a data analysis with preconceived notions and a hypothesis that you’re looking to prove. Let the data do the talking.
  2. Look for patterns, not isolated incidents. Avoid making sweeping inferences based on just a few data points.
  3. Confirm the trends you’re seeing are occurring elsewhere. Find supportive evidence beyond your own customer data to prevent false correlations. 

Big data’s effect on the traditional sales cycle

The traditional sales cycle has been altered, thanks to the growing access to customer data. Customers expect businesses to know what they want before they tell them, and at times, to introduce them to products they didn’t know they wanted. If you’re considering starting to leverage data insights for ads, marketing messages, direct mail, in-location stock, and more, the methodology outlined above should act as your North Star.

Are you a retailer? How are you using input from your customers to strengthen and optimize your sales messages? Share your thoughts in the comments below!