Leveraging data for Store specific decisions
Data is increasingly viewed as a very valuable asset. Many industry leaders call Data the new “Oil”. Several companies use data for CRM Analytics and targeted communication. However, do data assets have use beyond CRM? You betcha! Let’s look at few store and merchandising decisions today:
Understanding store performance
Sales data can be used for two broad types of store sales analysis:
1) Descriptive – What drives store sales? And where should I locate my store?
One can build a statistical model to understand which factors affect stores sales and to what degree - Catchment area Demographics, Market Potential, Cannibalization, Competition, Product / service offerings, and/or Customer satisfaction levels. These models can be used to select the right location for a new store
2) Predictive – What are store sales likely to be? And how well are stores performing?
One can forecast store sales based on – historical sales data, market events, competition, pricing changes, campaigns etc. This helps us track store performance versus actual sales-out. Allowing us to focus on stores that need our attention
Illustrative examples: Factors driving store performance, Focusing on the right stores based on their potential versus actual performance
We have helped clients understand store cannibalization. Just like a new product can cannibalize sales of its predecessor, new stores can also have overlapped catchment areas. Through traceable sales data, one can understand how many customers have started splitting spends (soft cannibalization) across stores or have permanently moved (hard cannibalization) to the new store.
Illustrative example: Measuring cannibalization from a new store
| City||Location ||%Hard Cannibalized ||% Soft Cannibalized |
| Mumbai||Goregaon ||4% ||67% |
| Mumbai||Malad ||3% ||57% |
Once cannibalization is established, the obvious next step is to quantify its financial impact and future steps
Stocking and Merchandising decisions:
It’s necessary to understand the demand-supply continuum for the product offerings. To begin with one needs to forecast demand (Details for another day!) and track sales out. This will enable taking timely pricing decisions if sales don’t track per plan. Similarly, this will allow timely restocking decision if sales exceeds forecast
Illustrative example: Building forecasts, Tracking sales out versus forecasts and inventory levels
Data can also help us stock the right locations. Thus minimizing stock outs at key locations and preventing excess inventory at the “not so key” locations.
Illustrative example: Ensuring appropriate stocking across different store types
But what should the retailer stock? Past sales data can give pointers into product adoption by different customer segments. Adoption along with understanding of segment growth can help us take precise merchandising decisions
Illustrative example: Understanding consumption patterns for different product types for customer segments and stores
Store layout decisions:
Based on past sales, store managers are now actively re-configuring store layout:
- Which departments should get greater focus?
- Which departments should be next to each other or far apart?
- Can I shelf and stock based on sales information to maximize revenue or profits?
- How can I customize my store layout to suite the store catchment area?
- Can I size my departments optimally to maximize store sales?
Information contained in the shopping basket can significantly aid above decisions – Which items were bought together, was the customer visit focused on certain items or did it fulfil a multitude of needs, which sections of the store did the customer visit, which ones are frequented more often than others, do products have seasonality
Illustration: Space allocation based on classification of items
Focus on location versus amount of space
Leverage location and amount of space
Use to pull customers in
Lastly, Retailers need to realize that increasing size of a department or section gives diminishing future returns. An optimization model can aid getting to the right size, adjacencies for a department.
Illustrative example: Space elasticity
In a nutshell:
Sales data at various levels of aggregation can really support store level decision making. From strategic decisions to helping solve tactical day-to-day problems. Don’t hesitate to explore if a given business problem can be solved better through available data and Analytics!