| Sales Analytics |
| |
Sales managers, who manage multiple products which are being sold across a large geographic market through multiple channels, spend huge resources to create a large data warehouse, which captures sales data to
the last level of detail and provides them with views of the data in multiple dimensions. These reports and trends help them to plan product and marketing initiatives for the future. But quite often despite having the most
updated sales data warehouse, they are unable to understand why sometimes their customer loyalty drops or their products do not form a part of the preferred customer marketing basket.
Typical problems faced by marketers include their inability to identify:
- declining customer loyalty
- customer segments/individual customers with irregular/infrequent
- purchase patterns to initiate appropriate sales push
- preferred collection of items typically purchased together by customers -
- this would help in designing newer and more effective promotions
- customers to target who purchase incomplete baskets - identification of customers who are not following the preferred baskets, and enable sales action
- impact of historical promotions on sales/margins
- segments of customers likely to respond to promotions
- impact of proposed new promotions
Historical sales data with product and customer information is the starting point for the data mining process, which can analyze the data to discover hidden patterns between the products or products and customers. These give valuable insights to the manager about frequent patterns in products sold together, customer behavior and loyalty, and directions for sales initiatives.
Further, if data on sale prices, product costs, past promotions, etc. are available then sophisticated data mining techniques can also reveal impact of current sales trends on margins, optimal mix for improving margins, impact of promotions customer segment-wise and plans for new promotions. These are just a few of the areas where data mining can help the mangers take better actionable decisions and improve the bottom line. |
| |
 |
| |
Data mining can help address these issues for improved sales performance:
In today's world, statisticians have been able to develop complex algorithms.
Standard tools are available for data mining experts to address these problems
and help managers to take decisions based on hard facts rather than on gut- feel.
A team of data analysts and business analysts take a multi stage approach to create a model using these sophisticated tools.
Steps include:
- understanding the customer business and data
- creating the data set to be used by the data mining tool
- create and validate the model using the tool
- run and fine tune the model
- set up the model for production run
Tools adopted by us have more than 20 sophisticated techniques including
neural networks, cluster analysis, classification, predictive methods, basket
analysis, etc. The final deliverables includes reports and charts which help sales
managers address their customers and market more effectively.
The Offshore Model
WE adopts one of two models for its Analytics practice
The hybrid model involves ICTEAS consultant meeting with the customer and working out the ETL model and building the list of deliverables. Subsequent analytics work happen offshore with frequent interaction with customer which is based on a time lines being maintained on intranet based project management and communication software. The customer is provided with relevant access privileges and has complete transparency on the progress of the project.
For smaller projects, a complete offshore model is used. Here again, the intranet project management and communication solution is used for monitoring the progress. |
