Maintaining Your Customers: Data Driven CRM

Customers today are very demanding and retailers must be able to anticipate customer needs or risk losing that customer.

For example: I grew up in a small town on the east coast of the United States. Every year before school started my mother would take me school shopping at the two or three local specialty stores in town. We knew the store owners by name and the owners knew our families. There was a sense of loyalty and the relationship between business and customer was a long standing one. If we did not find what we wanted or if we needed another size, the store clerk could place a special order and we waited, sometimes two weeks, for the product.

Today, back to school shopping is a very different process. We may go to one or two national retailers. We rarely know the store owners. Many times we shop online because it is possible to have access to regional, national or global retailers’ right in the privacy of our own home. Even though we do not know the merchants personally, we expect the retailers to understand and meet our needs. If we have to wait or if the retailer can not provide personalized customer service we will move on to the next merchant or supplier.

“We as customers want what we want and we want it now and at the right price.”

Fortunately today we have analytical tools and customer and product data to better serve and personalize the retail experience. Businesses know that it is less expensive to maintain a customer than to find a new customer. Therefore, customer relationship management (CRM) is key to a successful retail business. One CRM strategy that is gaining popularity is data mining.

Data mining is a process of knowledge discovery that employs statistical modeling to large amounts of transformed transactional and historical data, in an effort to understand and predict customer behavior.

The data utilized in the data mining process is typically massive and transactional in nature. It is opportunistic. The data was not collected with analysis in mind. Instead it was collected to monitor process control and track inventory. The data is usually messy and lacking integrity.


Right now ask yourself the following questions: Do you have access to large amounts of data? Is this data operational and/or transactional in nature?

If the answer is yes, you are lucky. You have the data needed to better understand your customers, hence making you a more responsive business partner. By uncovering patterns and relationships among your current customers, you will be able to put promotions and advertising in place that speaks directly to those customer segments. Once the power of your transactional data is unleashed you will be able to predict when a customer is about to churn. This will allow you to be proactive in the market place. Furthermore, you will be able to cross sell products and up sell products to existing customers.

You have the data what’s next? What is next is a process of data cleaning and data warehousing. You need to organize your transactional data so that you can build predictive models. The diagram below outlines the process. Stay tuned for the next edition of this newsletter when I will tell you how to household and prepare your data for the data mining process.