"Data mining as a tool for building and managing
Better Customer Relationships"


By

Vijay Prakash Anand
Faculty, Marketing
Praveen Kumar
Faculty, IT and System
IES Management College and Research Centre
'Vishwakarma' M.D. Lotlikar Vidya Sankul, Bandra Reclamation, Mumbai-400 050
 



1. Introduction

In the era of cut throat competition globally as well as locally, Data Mining is used by almost all the proactive corporations and organizations to build and manage customer relationships. The saying in Marketing - 'Customer is the King' has now been made possible by using the data mining applications by not only maintaining a good relationship with the customers; but also by effectively utilizing the data warehouses to add new customers.

Data Mining helps to retain the customers by understanding and fulfilling their needs proactively and thus delighting them in the long run. Earlier it was very difficult to understand and manage the data. But now with the usage of CRM Software from companies like SAP, Siebel, Oracle, Amdocs and others; many companies are utilizing the benefits of enhancing customer loyalty through Customer Relationship Management. In the banking sector in India, New Age Private banks like ICICI Bank, HDFC Bank, Axis Bank are at the forefront of utilizing the data mining techniques to enhance the customer relationship.

2. Data Mining

Data mining is the principle of sorting through large amounts of data and picking out relevant information. It is usually used by business intelligence organizations, and financial analysts, but it is increasingly used in the sciences to extract information from the enormous data sets generated by modern experimental and observational methods.

The Internet and technology opens up a wealth of information 24 hours a day, seven days a week, thereby heightening the transparency of the markets. Customers use the Internet to quickly shop around and see what competitors can provide. As a consequence, the attention span of customers has decreased, and customer loyalty is subject to new laws. Customers are looking beyond products to assess whether the overall solution you provide addresses their individual needs and priorities. Customers and knowledge about customers is one of most important assets of today's organization. The period of time between a new customer request and its fulfillment is decreasing. This means that, if one doesn't react quickly enough, the customers will find someone else.

Technology has also paved the way for a new dimension of customer relationship management. The falling costs for computing power and the arrival of new software tools for capturing and analyzing mass data have provided the main thrust behind the increase in importance of analytical solutions in general. Powerful hardware and software give better ways than ever before to understand and leverage customer relationships. Data mining is one of the technologies which provide analytical ability to the organizations for leveraging on customer relationships and thus customer loyalty and this paper analyzes the potential of data mining for building and managing better relationships. (META, 2007)

Data Mining uses a variety of techniques to find hidden patterns and relationships in large pools of data and infer rules from them that can be used to predict future behavior and guide decision making. (Fayyad et al., 2002; Hirji, 2001) To use data mining effectively for managing customer relationships the data must be categorized in some manner if it is to be accessed, re-used, organized, or synthesized to build a picture of the organization's competitive environment or solve a specific business problem (Pearlson, 2001, p.196). In recent years, the need to extract knowledge automatically from very large databases has grown.

Customer relationship Management form a learning relationships with the customers by noticing their needs with the use of online transaction processing i.e. operational CRM, remember their preferences with the use of Decision support Data warehouses and learn how to serve them with the use of Data mining. There may be number of channels by which company interfaces with its customers for example direct mail, Email, telemarketing etc. And key data for customer analytics come from these channels only and CRM is about serving customers through all these channels. Analytical CRM (Laudon, 2006) includes applications that analyze customer data generated by operational CRM applications to provide information for improving business performance management. For example developing customer profiles, analyzing customer profitability etc.

Data Mining is defined as exploration and analysis of large quantities of data by automatic or semi-automatic means to discover meaningful patterns and rules and these patterns allow a company to better understand its customers, improve its marketing, sales, and customer support operations. (Berry and Linoff, 1997)The process of extracting hidden key information from a large pool of available data. Data mining is often done to analyze data in order to gain knowledge about the behavior patterns of customers and to identify key relationships that may help in decision making.

Advanced statistical tools are used in data mining to understand current behavior and to predict future behavioral patterns. Mathematics, genetics, cybernetics and other fields of research make extensive use of data mining. In CRM, Web mining (pertaining to Web-related information) is used to gain insights into customer behavior. Interesting little poll over on KD Nuggets today - readers were asked where they had applied data mining in the last 12 months. The top 5 were CRM (26.1% of respondents), Banking (23.9%), Direct Marketing/ Fundraising (20.3%), Science (18.8%) and Fraud Detection (18.8%).

3. Customer Relationship Management

CRM can be defined as the process of predicting customer behavior and selecting actions to influence that behavior to benefit the company (Jenkins, 1999), usually leveraging on information technology and database-related tools.

CRM initiatives usually seek to fulfill several objectives. One of the objectives is to get closer to the customer by utilizing the data "hidden" in scattered enterprise databases. Examining and analyzing the data can turn raw data into valuable information about customer's needs. By predicting customer needs in advance, businesses can then market the right products to the right segments at the right time through the right delivery channels. Customer satisfaction can also be improved through more effective marketing.

Another objective of the CRM initiative is to transform the company into customer-centric organizations with a greater focus on customer profitability as compared to line profitability. The insights gained from CRM enable companies to calculate or estimate the profitability of individual accounts. Other CRM objectives include increased cross-selling possibilities, better lead management, better customer response and improved customer loyalty (Chin, 2000).

Analytical CRM

Analytical CRM aims at storing, analyzing and applying the knowledge about ways to approach customers, typically using data mining. Analyzing customer relationships from a lifetime perspective is critical for success.

The Data Monitors report titled "Analytical CRM," forecasts that global enterprise investment in analytical CRM will grow from an estimated $2.3 billion today to more than $3 billion in 2009. By employing analytical CRM analytics, businesses stand to gain a fuller understanding of their customers in order to serve them better, thus increasing customer longevity and generating more profit. Analytical CRM is the active collection, concentration and analysis of data gathered about the customer and his interactions with the business.


Figure 1: Customer Life Cycle

3.1 Widening the relationship with customers by acquiring new and profitable customers

To widen your customer relationships, Data mining can answer questions like:

Which kind of customers would you like to acquire?
Which kind of customers will drive your growth in future?
Which new customers are likely to be interested in your products?

Customer acquisition is the number one issue for small company marketers. As corporations increase budgets to attract and obtain new customers, data mining becomes a critical tool for profiling good customers, performing market segmentation, and improving the results of direct-marketing campaigns. The number of campaigns that can be managed in a given time period is often much lower than what the business demands. Data mining solves these problems by putting tools in the hands of the marketers driving these campaigns -- it keeps control where the need is. As a result, marketers can be much more responsive to creating new campaigns and can implement a direct feedback loop to improve their efforts on a regular basis.

3.2 Lengthening the relationship with your top customers by targeting existing  resources and strengthening the foundation of those relationships

To lengthen your customer relationships, Data mining can answer these questions:

Which customers in particular do you want to keep?
Which customers will drive most of your profits?
Which customers might switch to your competitors and why?
Which customers are dissatisfied with your services and products?

Customer retention: This retention is also a major issue for all businesses. One Harvard study suggests: "Reducing customer attrition by 5 per cent can double a company's profits." Given the high cost of finding new customers, a key issue for many organisations is customer retention. Often referred to as chum, customer turnover is a difficult problem to manage because it usually occurs without warning. For example, when a customer calls their long-distance carrier to have their account closed in favour of a competitor, the telecommunications provider knows only at that moment that their valued customer is churning. Once they are predisposed to leave, it is unlikely that the customer can be convinced to stay.

Data mining introduces a major paradigm shift to churn management by adding predictive capabilities. Data-mining tools can be used to model the patterns of past churning customers by examining billing histories, demographic information, and other customer data. Then, the same model can be used to predict other good customers who are likely to leave in the near future. Armed with this information, the marketer can proactively instigate campaigns to keep their customer, rather than fighting to get them back later.

3.3 Deepening the relationship with customers by transforming minor customers into highly profitable ones

Intensifying and deepening customer relationships also require Data mining tools to answer essential questions, such as:

With which customers can you increase the share of wallet?
Which products and services interest a particular customer?
Which products are typically bought together? Which crossselling opportunities should you consider?

Cross-selling: Growing a customer's value is yet another critical marketing function. The notion of increasing customer share is key to most organisations. Unlike increasing market share, which focuses on obtaining a greater number of customers, increasing customer share refers to getting more of the dollars each individual customer has to spend. Two common methods for this are customer-based product-launch campaigns, and cross-selling.

Riddled with as much guesswork and gut instinct as they are today, these methods are often not as effective as they could be. Data-mining tools improve product launches to an installed base, as well as cross-selling activities by helping marketers understand which customers are most likely to purchase new products, and which products are typically purchased together. This results in a more focused effort to customers ready to spend additional dollars.

Personalization provides relevant and specific recommendations for individuals, taking into account personal preferences, demographics, and behavior. Personalization permits delivering recommendations with the touch and timing of someone who knows you well.

Personalization uses data mining technology to analyze the large amounts of data gathered from Web sites and other applications to find patterns within purchase, demographic, ratings, and navigational data.

Personalization makes recommendations using data mining technology without the need to explicitly define manual rules. Personalization automatically deduces the customer's interests from the customer's behavior.

Personalization collects customer profile data and uses them to build predictive models that support personalized recommendations. The underlying rules derived through data mining can be more sophisticated and thus yield better results than the other techniques noted above. For example, "a person who has clicked links x and y and who has demographic characteristics a and b is likely to buy z ."

We have explored the CRM analytics with the following example of ICICI Bank, where they use data mining as tool for building and managing customer relationship.

In India, ICICI Bank is using the techniques of data mining to acquire new customers. These customers may be totally new to ICICI or in most of the cases; ICICI Bank taps their customer base and offers them various other services. For example: If you are having an ICICI Bank Account, you have special privilege offers from the bank for new services like Home Loans, Car Loans, Credit Cards, Personal Loans and others. ICICI Bank has also got Loan on Phone Scheme and Pre Approved Offers into each of their services; so that the relationship with the customer can not only be maintained, but will also make them more loyal of offering new services. Ultimately in the long run, the business of ICICI Bank will grow manifold by using the data mining techniques.

4. Data mining as a tool for CRM

The first analytical step in data mining data description– for example, summary of statistical attributes (such as means and standard deviations), may represented by using charts and graphs, and look at the distribution of values of the fields in your data.

One needs to build a predictive model based on patterns determined from known results, and then test that model on results outside the original sample. A good model should never be confused with reality, but it can be a useful guide to understanding the business. 

Data mining can be used for both classification and regression problems. In classification problems you're predicting what category something will fall into-for example, whether a person will be a good credit risk or not, or which of several offers someone is most likely to accept. In regression problems is like predicting a number such as the probability that a person will respond to an offer.

In CRM, data mining is frequently used to assign a score to a particular customer or prospect indicating the likelihood that the individual will behave in the way organization wants. For example, a score could measure the propensity to respond to a particular offer or to switch to a competitor's product. It is also frequently used to identify a set of attributes (called a profile) that segments customers into groups with similar behaviors, such as buying a particular product.

A special type of classification can recommend items based on similar interests held by groups of customers. This is sometimes called collaborative filtering.

There are some basic steps of data mining for CRM analytics:

1. Defining the business problem. Each CRM application will have one or more business objectives  for which you will need to build the appropriate model. Depending on your specific goal, such as "increasing the response rate" or "increasing the value of a response," you will build a very different model. An effective statement of the problem will include a way of measuring the results of your CRM project.

2. Building a marketing database. To build the marketing database, the data cleaning up process is required for building good models. The data needed may reside in multiple data resources such as the customer database, product database, and usage databases. This means integrate and consolidate the data into a single marketing database and reconcile differences in data values from the heterogeneous sources. There are often large differences in the way data is defined and used in different databases. Some inconsistencies may be easy to uncover, such as different addresses for the same customer. Making it more difficult to resolve these problems is that they are often subtle. For example, the same customer may have different names or multiple customer identification numbers may exist.

3. Exploring the data. Before building good predictive models, one must understand the consolidated data. Graphing and visualization tools are a vital aid in data preparation, and their importance to effective data analysis cannot be overemphasized. Data visualization most often provides good insights. Some of the common and very useful graphical displays of data are histograms or box plots that display distributions of values. Some may also want to look at scatter plots in two or three dimensions of different pairs of variables. The ability to add a third, overlay variable greatly increases the usefulness of some types of graphs.

4. Preparing data for modeling. This is the final data preparation step before building models and the step where the most "art" comes in. There are four main parts to this step:

First select the variables on which to build the model. Ideally, take all the variables, feed them to the data mining tool and let it find those which are the best predictors.

The next step is to construct new predictors derived from the raw data. For example, forecasting credit risk using a debt-to income ratio rather than just debt and income as predictor variables may yield more accurate results that are also easier to understand.

Next decide to select a subset or sample of data to build predictive models. If there is lot of data, however, using all data may take too long or require buying a bigger computer than you would like. Working with a properly selected random sample usually results in no loss of information for most CRM problems. Given a choice of either investigating a few models built on all the data or investigating more models built on a sample, the latter approach will usually help to develop a more accurate and robust model of the problem.

Last, transform variables in accordance with the requirements of the algorithm for building the model.

5. Building Data mining model. The most important thing to remember about model building is that it is an iterative process. Explore alternative models to find the one that is most useful in solving business problem.

Most CRM applications are based on a protocol called supervised learning. Start with customer information for which the desired outcome is already known. For example, use historical data because previously mailed to a list very similar to the contemporary one. Or conduct a test mailing to determine how people will respond to an offer. Then split this data into two groups. On the first group train or estimate the model. Then test it on the remainder of the data. A model is built when the cycle of training and testing is completed.

6. Evaluating results

Perhaps the most overrated metric for evaluating results is accuracy. Suppose an offer to which only 1% of the people will respond. A model that predicts "nobody will respond" is 99% accurate and 100% useless. Another measure that is frequently used is lift. Lift measure the improvement achieved by a predictive model. However, lift does not take into account cost and revenue, so it is often preferable to look at profit or ROI.

7. Incorporating data mining in CRM solution

In building a CRM application, data mining is often only a small, albeit critical, part of the product. For example, predictive patterns through data mining may be combined with the knowledge of domain experts and incorporated in a large application used by many different kinds of people.

The way data mining is actually built into the application is determined by the nature of the customer interaction. There are two main ways of interactions with the customers' i.e. inbound or outbound interactions. The deployment requirements are quite different.



Outbound interactions are characterized by our company originating the contact such as in a direct mail campaign. Thus selecting the people to whom you mail by applying the model to customer database. Another type of outbound campaign is an advertising campaign. In this case you would match the profiles of good prospects shown by your model to the profile of the people your advertisement would reach.

For inbound transactions, such as a telephone order, an Internet order, or a customer service call, the application must respond in real time. Therefore the data mining model is embedded in the application and actively recommends an action.

In either case, one of the key issues you must deal with in applying a model to new data is the transformations you used in building the model. Thus if the input data contains age, income, and gender fields, but the model requires the age-to-income ratio and gender has been changed into two binary variables, you must transform your input data accordingly.

5. Conclusion

Thus Customer Relationship Management is essential to compete effectively in today's marketplace. The more effectively you can use the information about your customers to meet their needs the more profitable you will be. Operational CRM with the help of analytical CRM with predictive data mining models as its core provides the business a cutting edge to maintain and enhance relationship. The path of a successful business like ICICI Bank and innumerable other corporations and organizations is build on data mining as the guiding factor.

References

1. META Group, "To CRM and Beyond", Research Note ADS 805, 2007

2. Fayyad U., Grinstein G.E. & Wierse A. (2002). Information Visualization in Data Mining and Knowledge Discovery. Morgan Kaufmann Publishers. Academic Press.

3. Pearlson, K.E. Managing and Using Information Systems: A Strategic Approach. New York, Wiley, 2001: John Wiley & Sons, Inc.

4. Laudon, K.C. and Laudon, J.P. Management information System: Managing The Digital Firm, Pearson, 2006.

5. Berry M.J.A. & Linoff G. (1999) Mastering Data Mining: The Art and Science of Customer Relationship Management. John Wiley and Sons, Inc.

6. http://www.insidecrm.com/dictionary/data-mining/

7. http://www.kdnuggets.com/polls/2007/data_mining_applications.htm

8. Jenkins D (1999). "Customer relationship management and the data warehouse". Call Center Solutions, 18(2): 88-92.

9. Chin J (2000). "It's important to do it well". Straits Times-Computer Times, 8 Nov, 2000: 14-16.

10. http://www.computerworld.com/softwaretopics/crm/story/ August 10, 2005 (Computing South Africa) JOHANNESBURG, South Africa
Datamonitor

11. Groth Robert www.thehindubusinessline.com/2000/12/11/stories/211177wp.htm, "Deriving meaning through mining"

12. http://www.sap.com/solutions/businesssuite/crm/pdf/AnalyticalCRM_50046585.pdf

13. Two Crows: Data Mining Glossary. Retrieved April 28, 2002 from: http://www.twocrows.com/glossary.htm
 


Vijay Prakash Anand
Faculty, Marketing
Praveen Kumar
Faculty, IT and System
IES Management College and Research Centre
'Vishwakarma' M.D. Lotlikar Vidya Sankul, Bandra Reclamation, Mumbai-400 050
 

Source: E-mail April 21, 2008

          

Articles No. 1-99 / Articles No. 100-199 / Articles No. 200-299 / Articles No. 300-399 /
Articles No. 400-499 / Articles No. 500-599 / Articles No. 600-699 / Back to Articles 700 Onward
Faculty Column Main Page