Data Mining and Customer Relationship Management
Customer Relationship Management (CRM) is important for every business. With the help of CRM, companies can satisfy their customers and increase their sales critically. As a broad aspect of a business, a number of methodologies and approaches are used to increase CRM capabilities. I want to briefly discuss a number of data mining techniques commonly used by CRM applications. This is an introductory post to the topic.
Customer Relationship Management
Throughout the history, many businesses have practiced various methods on their customers to realize their goals such as profit, customer satisfaction and loyalty. With the emergence of Information Technology, computers have been used in this field too.
Today, companies start interacting with their customers through different mediums with the help of different people and departments such as sales, marketing and maintenance. The amount of information stored about a customer increases day by day and duplication, impurity and most importantly unity of data are questionable in many ways. Overcoming these problems and controlling the customer relationship through one organized software system, not only solves these problems but also creates a brand new opportunity: understanding the patterns and knowledge hidden in this data for a specific goal.
Although the term, CRM, is mostly used for the software system managing the relationship, it has a broader meaning: a comprehensive strategy and process of acquiring, retaining and partnering with selective customers to create superior value for the company and the customer.
From technical point of view from Xu et. Al., CRM is about methodologies, software and internet. Companies can “build a database about its customers that depicts relationships in sufficient detail so that management, sales people, people providing service, and perhaps the customer directly, could access information, match customer need with product plans and offerings, remind customers of service requirements, know what other products a customer had purchased, and so forth.”
Data Mining
Data mining is a sophisticated data search capability that uses algorithms to discover insight and hidden information: patterns and correlations. It has a complementary approach to other techniques such as statistics and online analytical processing and plays an important role in decision making process by revealing unknown information.
Discovering such actionable knowledge has a key role in the facilitation of intelligent decisions and ability to create value. Among its vast amount of techniques and use cases, data mining techniques can be grouped into two according to the goal:
- Supervised Models: In these models, the effects of inputs on outputs are analyzed and event prediction or numeric value estimation can be achieved. Event prediction is classifying cases into a number of predefined groups. In numeric value estimation, value of a continuous field is estimated based on the observed input values.
- Unsupervised Models: In these models, there are no specific target outputs; instead the target is uncovered data patterns from the input. These models can analyze the input patterns and come up with the natural groupings with the help of Cluster models. Another use case of these models is the detection of associations between discrete fields and detecting associations over time.
Data Mining and Customer Relationship Management
As discussed above, data mining and CRM are concepts from two different but interconnected areas with different granularity and focus. While CRM is mostly about the management strategy of customer data for a goal like higher profit, data mining is a field where companies make use of computers for a goal.
Since data mining is a way to reveal insights from data, it is frequently used by CRM applications. A number of data mining techniques and their usages in this field:
- Clustering: This technique is used for segmentation of customers and defining target groups easily among millions of customers. As a result of clustering, one gets a high-level view of the database and understands what is going on.
- Classification & Prediction: These are data analysis techniques which are commonly used to describe important data classes or to predict future trends and numbers. Prediction of future customer behavior is an important process which allows companies to predict the trends of future customers and act accordingly.
- Association Rule Mining: This technique is used for recommendation. The insights about the customer interests may increase the cross-selling activities by recommending new items accordingly. The outputs reveal the preferences of customers and enables an understanding: which associated products a customer is more likely to buy or which types of campaigns customer is more likely to respond. An important example for this is the famous Market Basket Analysis and Product Placement Strategy or simply a movie recommendation system by Netflix. Similarly, HepsiBurada also uses a similar method for their customers.
- Combined Approach: Although above techniques may be utilized for different cases independently, they can be used in conjunction and combination with each other to realize better outcomes. A company can make use of clustering to define its target customers for a campaign such as customers with low response time. Then, classification and prediction techniques can be applied to determine the trends among this group. What are the purchase trends in this group? Finally, with Association Rule Mining, these trends are recommended to probable buyers. As a result, company can observe higher sales revenue by simply combining different techniques.
References
- Rygielski, Chris, Jyun-Cheng Wang, and David C. Yen. "Data mining techniques for customer relationship management." Technology in society 24, no. 4 (2002): 483-502.
- Kadiyala, Savitha S., and Alok Srivastava. "Data Mining For Customer Relationship Management." International Business & Economics Research Journal (IBER) 1, no. 6 (2011)
- Tsiptsis, Konstantinos, and Antonios Chorianopoulos. Data mining techniques in CRM: inside customer segmentation. Wiley, 2011.
- Hosseini, Seyed Mohammad Seyed, Anahita Maleki, and Mohammad Reza Gholamian. "Cluster analysis using data mining approach to develop CRM methodology to assess the customer loyalty." Expert Systems with Applications 37, no. 7 (2010): 5259-5264
- Xu, Yurong, David C. Yen, Binshan Lin, and David C. Chou. "Adopting customer relationship management technology." Industrial management & data systems 102, no. 8 (2002): 442-452
- Farooqi, Md, and Khalid Raza. "A Comprehensive Study of CRM through Data Mining Techniques." arXiv preprint arXiv: 1205.1126 (2012)
May 2013