Retail Business Intelligence: Using Data for Better Decision Making

In retail like any other industry there are multitude of interacting business processes. Every business process has input and output data. Output data of one business process can be input data for another and vice versa. Hence theres a multitude of data and making sense of these data becomes a "haystack" problem. Most retailers analyze few of the important data of few important business processes to help them make better business decisions. However it might just happen that some key data might be ignored just due to the absence of arranged and prioritized data. Also it might just happen that the retailer is tracking a wrong KPI altogether due to an myopic approach of selection of data. To add to the problem retailers have disparate systems for different business processes. Collecting data from all these disparate systems and making sense out it is a difficult job.
The idea is that retailers should be focused on a very select data and select KPIs; but the most critical and correct ones. Hence the question to be asked is how does one know what are the correct and critical KPIs and data?
Retail business intelligence is a process of defining data and KPIs to enable retailers better streamline their decision making using data.
Retail business intelligence has two components the metadata and the KPI. Metadata gives definition and logical meaning to the data while KPI gives business meaning to the data. Data from all the disparate systems are pooled in a central location such as a data warehouse. Here the data is associated to metadata to arrange logically segregate and arrange them.
Next KPIs across business processes are defined. For example KPIs for distribution, logistic, category management etc are all defined in a centralized location. These KPIs then use the pooled data to derive a value of business importance.
The KPIs are also arranged based on the viewing authority. For example a Inventory Planning Head would be interested in KPIs such as DC inventory vs Sales, Target Inventory vs Actual Inventory etc while a Store Manager will be viewing KPIs such as Store Floor Space Utilization, Employee Utilization etc. Hence restricting user/ viewer based viewing of data helps the users to focus on their respective KPIs.
Centrally managing data and KPI helps retailers better maintain data and quickly derive sense out of it and hence in turn drive agility in better decision making.

Point of Sale...Not Just about Sales!!

Point of Sale tool is the front-end tool that a retailer uses to conduct sale transactions. It is used to create the transaction where in the sale associate scans items and then finally generates the print receipt for the customer and accepts tenders from the customer for the sold merchandise. But is that the only function that the point of sale system is expected to perform?

Point of Sale system being the only front end system for the retalier; can be put to use in lot many areas for the retailer to better execute his business.

Firstly, the point of sale tool could used to manage customer and perform analytics on cutomer information in order to serve the customer better .The business interaction between a customer and the retailer is done via the point of sale system. Hence capturing of customer information such as customer address, customer birthday could be done at the point of sale system. While making a transaction for the customer the transaction could be linked to the customer and this information could be persisted for analysis. The customers’ buying basket could be analyzed from such persisted data. The customer buying basket data could then in turn be used to create custom promotions for the customer.

Secondly, promotion is a crucial component for a retailer to increase footfalls in its shops. For a retailer that has shops spread across geography a given promotion might not succeed at all places. Instead of cascading a centrally defined promotion across all shops the retailer might execute local promotions via the ponit of sale tool itself.

Thirdly, the sale data captured via the point of sale system could be shared by the retailer with a supplier if the retailer enagages with the supplier in a VMI model. Instead of forecasting demand on the orders made by the retailer the supplier would now forecast demand based on the point of sale itself and hence the demand forecast would be more accurate as in doing so the “bull whip” effect is eliminated.