Retail Analytics - An Intro
Retail Industry as we all know is a high volume low margin business. I am talking about the Supermarket variety here. This industry has to cope with loads of data, especially with POS data, constantly being revised product list (their SKUs), stock availability data, commercial data to calculate margins, customer acceptance data (for substitution products), customer services data (returns, complaints and refunds), new customer data, etc.
Timely information should be available for various departments like: Finance, Operations, Supply Chain, Marketing, Customer Services, Analytics, Commercial, and Business Development. However, we all know that the information required for the above-mentioned departments will be different.
If you look into the IT area of the retailers, you will find people working frantically to delivery daily / weekly / monthly reports on time. Analysts keep very busy in mining the data and discover the hidden trends and insights. You will also find multiple data marts, information silos that are not integrated. Departments do not have visibility into the strategic plans and other initiatives, so that they can work united.
Recession has actually driven the sales up for the retailers. But despite that fact, this industry is bracing itself for higher productivity and the better preparedness for the testing times ahead.
The information required by the retailers can be categorised into the following categories:
1. Business Reporting: Budgeting, Planning, Forecasting, Actuals analysis, Transaction & Operations, SCOR Metrics Reporting, Campaign analysis, Complaints, Returns, Refunds, Sales, Margins etc.,
2. Business Analytics: Customer segmentation, Value Chain Analysis, Market Basket Analysis, Price Optimisation, Locational intelligence
Here Budgeting, Planning and Forecasting can be categorised under Performance Management, but here the focus is on Business Analytics.
Agenda behind Retail Analytics is: to intelligently cut costs, target customers better, and increase the profits.
Brief explanation on the Business Analytics' space:
Customer Segmentation: Retail customers are categorised into multiple segments based on their purchasing habits, and demographics. Clustering is done for different product groups to understand the customers better and target them for more wallet share. Most importantly, clustering involves store-wise / location-wise customer behaviour. These are very useful when introducing new products or brand extensions. To understand the changing customer behaviour across time and products, the data is made available in an interactive decision making (IDM) mode. Powerful OLAP engines will be very handy here. Data can be assembled in OLAP fashion after extensive transformation and aggregation of data. IDM also reduces the need for more reports.
Value Chain Analysis: Value Chain Analysis (VCA) tries to analyse products' journey (product movement from goods receipt to customer pickup) in terms of money spent. VCA aims to bring up understanding on packaging, weight carried, and shelf space occupied while at distribution centre / at store / on vehicle - all in terms of money spent. It also tries to match the shelf occupancy space compared to the customer pickup rate, to ensure that the product is optimally stocked at the back stores / freezer / front store and the stock order quantity. Economic Order Quantity (EOQ) has a new definition with the introduction of VCA. It enables the retailers to try out new methods without compromising the customer service and product availability. It helps the retailers to communicate well with the suppliers on the pack sizes, SKU sizes and the order quantities.
Market Basket Analysis: This analyses product combinations customers are likely to buy. Understanding of this improves the cross/up selling and marketing effectiveness for the retailers through product positioning, promotion offers, optimised product range, floor plans etc.,
Price Optimisation: This is usually employed by the retailers to improve their margins on best selling products, cut losses on slow-movers, and to keep up to the competition. Here the price elasticity is tested for margin improvements.
Locational Intelligence: LI enables the retailers to understand the geographical clusters of the customers who shop at their stores. This enables them to profile them and also to decide on location of their future stores. LI improves the readability when the clusters are combined with geographical maps.
