Professional Documents
Culture Documents
by
Nesibe ŞAHİN
Ahmet ÖNCE
June 2004
A CRM APPLICATION IN GSM SECTOR
by
Nesibe ŞAHİN
Ahmet ÖNCE
A thesis submitted to
of
Fatih University
June 2004
Istanbul, Turkey
ii
A CRM APPLICATION IN GSM SECTOR
Nesibe ŞAHİN
Ahmet ÖNCE
ABSTRACT
CRM and applications of CRM in some sectors are examined. Information about
data analysis and data mining techniques are gathered. Simulated three-month-data of
calling times of a GSM operator is evaluated, most used tariffs according to Pareto
analysis are determined and focused on these tariffs. Statistical distributions of calling
times of these tariffs and central tendency values are determined. These three month
data is compared in between. After that, the cities that use this GSM operator mostly are
considered. The distributions of calling times of these cities’ most used tariffs are
determined. These data is evaluated and reports that cover all of these evaluations are
prepared. According to these reports, existing conditions are determined, problem in the
current system are defined and proposals to improve this system are done.
iii
GSM SEKTÖRÜNDE BİR CRM UYGULAMASI
Nesibe ŞAHİN
Ahmet ÖNCE
ÖZ
iv
TO MY FATHER, MOTHER AND MY LITTLE BROTHERS
(NESİBE ŞAHİN)
(AHMET ÖNCE)
v
ACKNOWLEDGEMENT
vi
TABLE OF CONTENTS
ABSTRACT.....................................................................................................................iii
ÖZ.....................................................................................................................................iv
DEDICATION…………………………………………………………………………vii
ACKNOWLEDGEMENT…………………………………………………………….viii
LIST OF TABLES……………………………………………………………………....x
LIST OF FIGURES……………………………………………………………………..xi
CHAPTER 1 INTRODUCTION...................................................................................1
1.1 CUSTOMER RELATIONSHIP MANAGEMENT (CRM)...................................1
CHAPTER 2 CUSTOMER RELATIONSHIP MANAGEMENT...............................3
2.1 HISTORY OF CRM MARKET..............................................................................3
2.1.1 Major Vendors..................................................................................................5
2.2 MISCELLANEOUS DEFINITIONS OF CRM......................................................6
2.3 DRIVERS FOR CRM APPLICATION..................................................................9
2.3.1 Reasons For Adopting CRM: The Business Drivers........................................9
2.3.2 Cost Goals......................................................................................................10
2.4 THE CRM INDUSTRY........................................................................................11
2.4.1 Size Of The CRM Industry.............................................................................11
2.4.2 Vendors...........................................................................................................12
2.5 INFORMATION TECHNOLOGIES FOR CRM.................................................13
2.5.1 IT Factors of CRM.........................................................................................14
2.6 RETURN ON INVESTMENT OF IMPLEMENTATION...................................15
2.6.1 Cost And Time...............................................................................................15
2.6.2 Benefits...........................................................................................................16
2.6.3 CRM: Commitment To Customer & Shareholder Value...............................17
2.6.4 ROI Of CRM Project......................................................................................18
2.7 PRINCIPLES OF CRM.........................................................................................18
2.8 CRM ISSUES........................................................................................................19
2.8.1 Customer Privacy...........................................................................................19
vii
2.8.2 Technical Immaturity.....................................................................................20
2.9 CASE STUDIES...................................................................................................21
2.9.1 Amazon.Com..................................................................................................21
2.9.2 Dell.................................................................................................................21
2.9.3 Volkswagen....................................................................................................22
2.9.4 Wells Fargo....................................................................................................23
CHAPTER 3 DATA MINING.....................................................................................25
3.1 What Is Data Mining?...........................................................................................25
3.1.1 Overview Of Data Mining..............................................................................26
3.1.2 Semma............................................................................................................27
3.1.3 Business Intelligence Using Data Mining......................................................30
3.2 PRIVACY.............................................................................................................31
3.2.1 OECD Principles of Data Collection..............................................................31
CHAPTER 4 CASE STUDY.......................................................................................33
4.1 Introduction...........................................................................................................33
4.2 Steps Of Our Work................................................................................................33
4.2.1 STEP 1: Determining The Most Used Tariffs And The Cities That The
Operator Is Used Most.............................................................................................33
4.2.2 STEP 2: Finding Statistical Values Of The Calling Times............................33
4.2.3 STEP 3: Preparing Reports of These Data.....................................................36
4.2.4 STEP 4: Evaluate These Reports....................................................................38
4.2.5 STEP 5: Tariff Recommendations For The GSM Operator...........................40
CHAPTER 5 CONCLUSION.....................................................................................42
5.1 A General Look To Crm........................................................................................42
5.2 Results Of Case Study...........................................................................................42
5.2.1 Identification Of Problems............................................................................42
5.2.2 Proposals To These Problems........................................................................43
APPENDIX A………………………………………………………………………….44
APPENDIX B………………………………………………………………………….45
APPENDIX C………………………………………………………………………….93
viii
APPENDIX D……………………………………………………………………….100
APPENDIX E………………………………………………………………………..103
REFERENCES………………………………………………………………………104
ix
LIST OF TABLES
TABLES
x
LIST OF FIGURES
FIGURES
xi
CHAPTER 1
INTRODUCTION
Before the advent of the supermarket, the mall, and the automobile, people went
to their neighborhood general store to purchase goods. The proprietor and the small
staff recognized the customer by name and knew the customer's preferences and wants.
The customer, in turn, remained loyal to the store and made repeated purchases. This
idyllic customer relationship disappeared as the nation grew, the population moved
from the farm communities to large urban areas, the consumer became mobile, and
supermarkets and department stores were established to achieve economies of scale
through mass marketing.
Although prices were lower and goods more uniform in quality, the relationship
between the customer and the merchant became nameless and faceless. The personal
relationship between merchant and customer became a thing of the past. As a result,
customers became fickle, moving to the supplier who provided the desired object at
lowest cost or with the most features.
The last several years saw the rise of Customer Relationship Management
(abbreviated CRM) as an important business approach. Its objective is to return to the
world of personal marketing. The concept itself is relatively simple. Rather than market
to a mass of people or firms, market to each customer individually. In this one-to-one
approach, information about a customer (e.g., previous purchases, needs, preferences
and wants) is used to frame offers that are more likely to be accepted. This approach is
made possible by advances in information technology.
1
Remember that CRM is an abbreviation for Customer Relationship Management,
not Customer Relationship Marketing. Management is a broader concept than marketing
because it covers marketing management, manufacturing management, human resource
management, service management, sales management, and research and development
management. Thus, CRM requires organizational and business level approaches –
which are customer centric – to doing business rather than a simple marketing strategy.
2
CHAPTER 2
Activity Management: Provide calendar and scheduling for individual sales people
Opportunity Management: Manage leads and potential leads for new customers
Order Management: Obtain online quotes and transform inquiries into orders
3
Document Management: Develop and retrieve standard and customizable
management reports and presentation documents
4
Today, CRM includes all customer-facing applications, including:
Sales Force Automation (SFA),
Customer Service (CS),
Sales and Marketing Management (SMM), and Contact & Activity
Management..
The major vendors changed over time. In 1993, the leaders of SFA were Brock
Control, Sales Technologies, and Aurum. Since then, Brock Control changed its name
to Firstwave Technologies, Inc. In 1998, Sales Technologies merged with Walsh
International and now is consolidated into SYNAVANT Inc. to provide pharmaceutical
and healthcare industry relationship management service. Aurum was merged into
Baan, which in turn was acquired by Invensys plc in July 2000.
In the CS area, Scopus, Vantive and Clarify were the major vendors. However,
things also changed rather rapidly:
Siebel merged with Scopus in 1995 and dominated the consolidated CRM
market with 68% market share.
In 1998, the CRM market was divided by Siebel, Vantive (now PeopleSoft),
Trilogy, and Clarify (now Nortel), and Oracle (in that order) plus fewer than 20 other
companies with small market shares.
At the beginning of 2000, Siebel Systems Inc. was the market leader with a 35%
share. Vantive (PeopleSoft) and Clarify Inc. (Nortel) followed. SAP and Oracle
Corporation were introducing new application to the market based on their software
development capabilities. Recent entrants offering Web applications and services
include Silknet Software, E.piphany, and netDialog.
5
2.2 MISCELLANEOUS DEFINITIONS OF CRM
“CRM is a business philosophy which provides a vision for the way your
company wants to deal with your customers. To deliver that vision, you need a CRM
strategy which gives shape to your sales, marketing, customer service and data analysis
activities. For most companies, the aim of a CRM strategy is to maximize profitable
relationships with customers by increasing the value of the relationship for both the
vendor and the customer.”
“It’s a business approach that builds customer loyalty and retention.” John
O’Connell, chairman and CEO of Staffware.
“Conquering barriers that prevent customers and companies from knowing each
other.” Margaret Gerstenkorn, business development associate at Oncontact Software.
6
“Our industry has not done a good enough job to make that value proposition
clear. I define CRM as a business approach that integrates PPT (people, process and
technology) to maximize relations with all customers. It’s not a one-off, but a complete
approach that coordinates customer-facing operations like sales, marketing, and
customer service. It should help sales, raise productivity, and improve employee
morale.” Barton Goldenberg, founder and president of ISM.
“CRM is a business strategy that goes beyond increasing transaction volume. Its
objectives are to increase profitability, revenue, and customer satisfaction. To achieve
CRM, a company wide set of tools, technologies, and procedures promote the
relationship with the customer to increase sales. Thus, CRM is primarily a strategic
business and process issue rather than a technical issue.”
Customer: The customer is the only source of the company’s present profit and
future growth. However, a good customer, who provides more profit with less resource,
is always scarce because customers are knowledgeable and the competition is fierce.
Sometimes it is difficult to distinguish who is the real customer because the buying
7
decision is frequently a collaborative activity among participants of the decision-making
process. Information technologies can provide the abilities to distinguish and manage
customers. CRM can be thought of as a marketing approach that is based on customer
information.
All of these techniques, processes and procedures are designed to promote and
facilitate the sales and marketing functions.
8
2.3 DRIVERS FOR CRM APPLICATION
In the past, the prime approach to attracting new customers was through media
and mail advertising about what the firm has to offer. This advertising approach is
scattershot, reaching many people including current customers and people who would
never become customers. For example, the typical response rate from a general mailing
is about 2%. Thus, mailing a million copies of an advertisement yields only 20,000
responses on average.
Another driver is the change introduced by electronic commerce. Rather than the
customer dealing with a salesperson either in a brick and mortar location or on the
phone, in electronic commerce the customer remains in front of their computer at home
or in the office. Thus, firms do not have the luxury of someone with sales skills to
convince the customer. Whereas normally it takes effort for the customer to move to a
9
competitor’s physical location or dial another 1-800 number, in electronic commerce
firms face an environment in which competitors are only a few clicks away.
10
2.4 THE CRM INDUSTRY
Estimates of the size of the CRM industry are shown in Table1 and plotted in
Figure 1. These illustrations show forecasts made in the 1997 to 2000 period by a
number of industry research groups. It is important to realize that the forecasters
generally did not specify what they included in their estimates. Therefore, it is not
possible to tell which expenditures (e.g., hardware, software, mailing, personnel, call
centers …) and which revenues are included in the numbers shown.
Not all values shown in Table 1 are forecasts; some of the values shown were obtained
by taking the forecaster’s growth rate and then interpolating. Interpolated values are
shaded.
11
Clearly the forecasts shown vary significantly as they reach the out years because
they are based on different assumptions of the size of the current market at the time of
the forecast and the growth rate inferred from the numbers presented. The important
point, is that the market is growing and is multibillion.
2.4.2 Vendors
A few years ago, technology vendors had their own specialties. For example,
Siebel was in sales force automation, Remedy was in helpdesk systems, Davox was in
call center systems, eGain was in e-mail management, and BroadVision was in the
front-end application area. Today, however, there is no specific boundary of vendors.
All vendors are trying to expand their products over the entire CRM area. For example,
Siebel says it can do everything, Davox moved into customer contact management, and
BroadVision is trying to integrate backward with ERP.
12
Table 2. 2: The Major CRM Vendors
CRM differs from the previous method of database marketing in that the database
marketing technique tried to sell more products to the customer for less cost. The
database marketing approach is highly company centric. However, customers were not
kept loyal by the discount programs and the one-time promotions that were used in the
database-marketing programs. Customer loyalty is, indeed, very difficult to obtain or
buy. The CRM approach is customer-centric. This approach focuses on the long-term
relationship with the customers by providing the customer benefits and values from the
customer’s point of view rather than based on what the company wants to sell.
13
Four basic tasks are required to achieve the basic goals of CRM.
1. Customer Identification
To serve or provide value to the customer, the company must know or identify the
customer through marketing channels, transactions, and interactions over time.
2. Customer Differentiation
Each customer has their own lifetime value from the company's point of view and each
customer imposes unique demands and requirements for the company.
3. Customer Interaction
Customer demands change over time. From a CRM perspective, the customer’s long-
term profitability and relationship to the company is important. Therefore, the company
needs to learn about the customer continually. Keeping track of customer behavior and
needs is an important task of a CRM program.
4. Customization / Personalization
“Treat each customer uniquely” is the motto of the entire CRM process. Through the
personalization process, the company can increase customer loyalty. Jeff Bezos, the
CEO of Amazon.com, said, “Our vision is that if we have 20 million customers, then we
should have 20 million stores.” The automation of personalization is being made
feasible by information technologies.
14
Table 2. 3: IT Factors in CRM
The 1999 Cap Gemini and IDC survey also found that, the average total
investment in CRM of 300 U.S. and Europe companies was $3.1 million. More than
69% of the companies surveyed spent less than $5 million, and more than 13% of the
companies spent over $10 million.
The cost of implementing a CRM system is easily double the Enterprise Resource
Planning (ERP) implementation cost. Average implementation time for an ERP system
is 23 months and the cost of ownership over the first 2 years is from 0.4% to 1.1% of
company revenue.
15
Table 2. 4: Annual CRM Expenses (in $million)
2.6.2 Benefits
16
Table 2. 6: Benefits of CRM project
More than 57% of CEOs in another survey with 191 respondents believe that the
major objective of CRM is customer satisfaction and retention. Another 17% said it is
designed to increase cross selling and up selling.
Ideally, CRM is a comprehensive strategy that integrates all areas of business that
touch the customer – though mainly, it is limited to marketing, sales, customer service
and field support — through the integration of people, process and technology. To be
successful, CRM requires acquiring and distributing knowledge about one’s customers
across the enterprise, to balance costs, revenue and profits with customer satisfaction.
Obviously, business processes and key technologies are required to optimize CRM
strategies.
17
In sum, CRM is four things that provide competitive advantage to the enterprise:
The overall processes and applications of CRM are based on the following basic
principles.
18
Treat Customer Individually,remember customers and treat them individually.
CRM is based on philosophy of personalization. Personalization means the
content and services to customer should be designed based on customer
preferences and behavior.’ Personalization creates convenience to the customer
and increases the cost of changing vendors.
Acquire and Retain Customer Loyalty through Personal Relationship
Once personalization takes place, a company needs to sustain relationships with
the customer. Continuous contacts with the customer – especially when designed
to meet customer preferences – can create customer loyalty.
Select “Good” Customer instead of “Bad” Customer based on Lifetime Value
Find and keep the right customers who generate the most profits. Through
differentiation, a company can allocate its limited resources to obtain better
returns. The best customers deserve the most customer care; the worst customers
should be dropped.
In summary, personalization, loyalty, and lifetime value are the main principles of
CRM implementation.
Customer privacy is an important issue in CRM. CRM deals with large amounts of
customer data through various touch points and communication channels. The
personalization process in CRM requires identification of each individual customer and
collections of demographic and behavioral data. Yet, it is the very information that most
customers consider personal and private. The individual firm is thus caught in an ethical
dilemma. It wants to collect as much information as possible about each customer to
further its sales, yet in doing so it treads at and beyond the bounds of personal privacy.
Privacy issues are not simple. There are overwhelming customer concerns, legal
regulations, and public policies around the world. Still it is unclear and undetermined
19
what extent of customer privacy should be protected and shouldn’t be used, but four
basic rules might be considered.
The customer should be notified their personal information is collected and will
be used for specific purposes.
The customer should be able to decline to be tracked.
The customer should be allowed to access their information and correct it.
Customer data should be protected from unauthorized usage.
Some companies provide ‘customer consent form’ to ask the customer to agree to
information collection and usage. Providing personalized service to customer is a way
to satisfy customers who provided their personal information. All of these efforts are
designed to build trust between the company and its customers.
The concept, technologies, and understanding of CRM are still in its early adapter stage.
Most of the CRM technologies are immature and the typical implementation costs and
time are long enough to frustrate potential users.
Many software and hardware vendors sell themselves as complete CRM solution
providers but there is little standardized technologies and protocols for CRM
implementation in the market. Even the scope and extent of ‘what CRM includes’ differ
from vendor to vendor; each has different implementation requirements to achieve the
customer’s expectations.
CRM is one of the busiest industries which occurs frequent merger and
acquisition. Many small companies merge together to compete with large vendor. Large
companies such as PeopleSoft acquired small vendor to enter this ‘hot’ CRM market.
Due to these frequent merger and acquisition, the stable technical support from the
market becomes rare. Vendors publish new version – maybe more integrated software –
of CRM software as frequently as they can and customers should pay for that.
20
Often these technical immaturities or unstable conditions are combined with the
customer requirements which are frequently unclear and lead the project failure. These
technical immaturities may be overcome over time, but the process may be long and
painful.
2.9.1 Amazon.Com
When you try to buy something from Amazon.com, you can see the following
statement; “Customers who bought this item also bought these items.” If you have any
previous purchasing experience with Amazon.com, the company will support a
‘Welcome to Recommendations’ Web page.
The personalized Web pages, vast selection of products, and low price lead
customer loyalty and long-term relationship of Amazon.com. More than 20 million
people have purchased at Amazon.com. The percentage of returning customers is about
15 to 25 percent, compared with 3 to 5 percent for other ebusiness retailers.
Like the corner merchant of old, Jeff Bezos, the founder of Amazon.com, believes
the Internet store of the future should be able to guess what the customer wants to buy
before the customer knows. He wants to make Amazon.com Web site that smart and
that personal.
2.9.2 Dell
Since 1983, Dell Computers has operated on two simple business ideas: sell
computers direct to individual customers and manufacture computers based on the
21
customer’s order. The individual customer can make his/her system unique and obtain it
directly from the company.
If the system has a problem, the user can contact the Dell Web site directly and get
personalized services by using the customer system service tag number, which is on the
side of the computer. These personalized services also provide related information and
make software downloads available. In addition, a call center provides technical
assistance at multiple levels. If the first level technician cannot resolve the problem, the
customer is routed to a more skilled contact.
2.9.3 Volkswagen
In 1988, the company started its ‘Customer Come First’ marketing strategy. Under
this strategy, all of the decision-making processes are based on the ‘Voice of Customer.’
The company carefully monitored their response to advertisements, customer
expectations, and customer satisfaction. Customer forums and focus group are used to
hear the customer voice.
22
The company maintains a central database to provide club card, bonus point
programs, club shops, and Volkswagen magazine. Every contact points with a customer
gives the company more information about the customer, so the company can constantly
improve the quality and value of the customer database.
Banking differs from other industries because the average relationship between
customer and bank lasts much longer on the. For example, in the auto industry, the
relationship between the customer and the company is becoming weaker over time. You
don’t need to contact the car dealer or manufacturer once a week or a month. You can
change your oil or maintain your car with different service station. However, once you
open your account in a specific bank, your relationship or dependence to the bank
increases. You may write checks more frequently, have direct deposit, transfer money,
pay bills, and withdraw money. The bank contacts you regularly by sending you your
monthly statement. You can obtain credit card or investment opportunities from the
bank.
Wells Fargo is one of the leading banks which transforms these relationships into
opportunities. It was the first bank which started 24-hour phone banking service and
opened branches in the local supermarket and Starbucks coffee house. Wells Fargo
always tried to provide more touch points to its customers and a one-stop shopping
environment.
Since 1993, Wells Fargo tried to integrate all of its back-end customer information
into its Customer Relationship System. Previously, customer information was managed
by several different backend system. Software was organized by account number, with
each backend system using its own numbering system. Customer service agents found it
difficult to integrate customer information when they received a request to transfer from
one account to another. They had to log on to several different system to obtain the
information and do the transactions requested. In the new system, the service agent can
access all required information by using the customer’s social security number instead
of the account numbers. These changes increase convenience for both customers and
service agents.
23
Wells Fargo provides Internet banking. It built a Web site as a new contact point
in 1995 and provided advanced technologies to its customer. By using online banking,
customers can manage their account anytime and anywhere. Online banking also saves
operating cost of the bank branches.
In the future, Wells Fargo will try to build online customer communities (similar
to America Online or the World Wide Web) in its banking service by responding to
customers’ needs with new technologies. By providing more power to manage their
account and money, Wells Fargo expects to increase customer loyalty and obtain long
term mutual benefits with its customers.
24
CHAPTER 3
DATA MINING
Data mining differs from traditional statistics in several ways: formal statistical
inference is assumption driven in the sense that a hypothesis is formed and validated
against the data. Data mining in contrast is discovery driven in the sense that patterns
and hypothesis are automatically extracted from data. Said another way, data mining is
data driven, while statistics is human driven. The branch of statistics that data mining
resembles most is exploratory data analysis, although this field, like most of the rest of
statistics, has been focused on data sets far smaller than most that are the target of data
mining researchers.
Data mining also differs from traditional statistics in that sometimes the goal is to
extract qualitative models which can easily be translated into logical rules or visual
representations; in this sense data mining is human centered and is sometimes coupled
with human-computer interfaces research.
Data mining is a step in the data mining process, which is an interactive, semi-
automated process which begins with raw data. Results of the data mining process may
be insights, rules, or predictive models.
The field of data mining draws upon several roots, including statistics, machine
learning, databases, and high performance computing.
25
3.1.1 Overview of Data Mining
To convert the value of the data warehouse or data mart into strategic business
information, many companies are turning to data mining, an emerging technology based
on a new generation of software. Data mining combines techniques including statistical
analysis, visualization, induction, and neural networks to explore large amounts of data
and discover relationships and patterns that shed light on business problems. In turn,
companies can use these findings for more profitable, proactive decision making and
competitive advantage.
Data mining was designed for exploiting massive amounts of data. This process
can be more efficient if you first define what the business problem is, and then
determine the amount of data you will need to solve the problem. By taking this
"bottom up" approach to data mining and involving upper management in the
understanding of business problems and the potential ROI, the process will be much
more acceptable and the goals attainable.
SAS Institute defines data mining as the process of selecting, exploring, and
modelling large amounts of data to uncover previously unknown patterns for a business
advantage. As a sophisticated decision support tool, data mining is a natural outgrowth
of a business investment in data warehousing. The data warehouse provides a stable,
easily accessible repository of information to support dynamic business intelligence
applications.
26
As the next step, organizations employ data mining to explore and model
relationships in the large amounts of data in the data warehouse. Without the pool of
validated and "scrubbed" data that a data warehouse provides, the data mining process
requires considerable additional effort to pre-process data. Although the data warehouse
is an ideal source of data for data mining activities, the Internet can also serve as a data
source. Companies can take data from the Internet, mine the data, and distribute the
findings and models throughout the company via an Intranet. Although data mining
tools have been around for many years, data mining became feasible in business only
after new hardware and software technology advances became available.
Software advances continued data mining's evolution. With the advent of the data
warehouse, companies could successfully analyze their massive databases as a coherent,
standardized whole. To exploit these vast stores of data in the data warehouse, new
exploratory and modeling tools--including data visualization, neural networks, and
decision trees--were developed. Finally, data mining incorporated these tools into a
systematic, iterative process.
3.1.2 Semma
Beginning with a statistically representative sample of the data, you can apply
exploratory statistical and visualization techniques, select and transform the most
significant predictive variables, model the variables to predict outcomes, and affirm the
model's accuracy. To clarify the data mining process, SAS Institute has mapped out an
27
overall plan for data mining. This step-by-step process is referred to by the acronym
SEMMA: sample, explore, modify, model, and assess.
Step 1: Sample
Extract a portion of a large data set big enough to contain the significant
information yet small enough to manipulate quickly. For optimal cost and performance,
SAS Institute advocates a sampling strategy, which applies a reliable, statistically
representative sample of the full detail data. Mining a representative sample instead of
the whole volume drastically reduces the processing time required to get crucial
business information. If general patterns appear in the data as a whole, these will be
traceable in a representative sample. If a niche is so tiny that it's not represented in a
sample and yet so important that it influences the big picture, it can be discovered using
summary methods.
Step 2: Explore
Step 3: Modify
Create, select, and transform the variables to focus the model construction process.
Based on your discoveries in the exploration phase, you may need to manipulate your
data to include information such as the grouping of customers and significant
subgroups, or to introduce new variables. You may also need to look for outliers and
reduce the number of variables, to narrow them down to the most significant ones. You
may also need to modify data when the "mined" data change. Because data mining is a
dynamic, iterative process, you can update data mining methods or models when new
information is available.
28
Step 4: Model
Step 5: Assess
Evaluate the usefulness and reliability of findings from the data mining process.
The final step in data mining is to assess the model to estimate how well it performs. A
common means of assessing a model is to apply it to a portion of data set aside during
the sampling stage sometimes known as validation data. For a model to be considered
successful and useful, it should work for this validation sample as well as for the
training data used to construct the model. Similarly, you can test the model against
known data. For example, if you know which customers in a file had high retention
rates and your model predicts retention, you can check to see whether the model selects
these customers accurately. In addition, practical applications of the model, such as
partial mailings in a direct mail campaign, help prove its validity.
The Future
29
mining techniques will be able to understand key business issues more thoroughly and
to present the results of analysis meaningfully to specialist marketing analysts and
general users alike. In learning more about themselves and their customers, these
organizations will see a shift towards true one-to-one relationships with the customers--
ensuring complete customer relationship management. Accurate anticipation of the
customers' actions can lead to increased effectiveness of marketing activities and
decreased financial risks.
Companies typically begin their business intelligence (BI) journey with a focus on
understanding and measuring the outcome of past decisions. But these “rear-view
mirror” technologies can’t provide you with a clear picture of the future — they only
give you a view of the road behind you. Industry leaders are realizing that forward
looking business intelligence is imperative to making better decisions that solve
business problems and keep their companies moving in a profitable direction. They’re
evolving their BI capabilities by adding data mining technology to their operations
because they know that if they don’t — they’ll perish at the hands of competitors that
do. Data mining looks forward to tell you what is most likely to happen — giving you
the power to improve your future. The most evolved business intelligence continually
applies data mining techniques and deploys the results enterprise-wide.
30
Figure 3. 2
This graph shows how a wireless telco has evolved their BI — an evolution to solving
the problems that affect future profits. They began with reporting that gave them simple
measurements. Added OLAP to drill-down to more detail. Focused their BI on the
future with data mining. And finally deployed data mining results to their front lines to
continually improve ROI.
3.2 PRIVACY
31
o Data quality: Data should be relevant to the stated
purposes, accurate, complete and up-to-date; proper precautions should be taken
to ensure this accuracy
o Purpose specification: The purposes for which
data will be used should be identified, and the data should be destroyed if it no
longer serves the given purpose
o Use limitation: Use of data for purposes other
than specified is forbidden, except with the consent of the data subject or by
authority of law
o Security safeguards: Agencies should establish
procedures to guard against loss, corruption, destruction, or misuse of data
o Openness: It must be possible to acquire
information about the collection, storage, and use of personal data
o Individual participation: The data subject has a
right to access and challenge the data related to him or her
o Accountability: A data controller should be
accountable for complying with measures giving effect to all these principles
32
CHAPTER 4
CASE STUDY
4.1 INTRODUCTION
4.2.1 STEP 1: Determining The Most Used Tariffs And The Cities That The
Operator Is Used Most
We used Pareto analysis to determine the most used and the most affective tariffs
from the simulated data. There were more than 50 tariffs in our simulated data. It was
impossible to evaluate all these tariffs so we determined the most affective ones which
are used totally more than %90 of customers.
33
Statgraphic is used as a statistic software and all the data are sent partially to the
Statgraphic in order to find statistical values. Menus that are used in our project will be
shown below:
4.2.2.1 Statgraphic
From this analysis, we used analysis summary and summary statistics tables. We
obtain statistical values of our data from these tables.
34
figure 4. 2: a sample of Analysis Summary window
35
4.2.2.1.2.1 Analysis Summary
36
We supported our reports with histograms, density trace and scatter plots. These
plots are obtained from one variable analysis and distribution fitting menus.
See Appendix B for report of detailed analysis of each tariff ( for month_x)
37
figure 4. 7: a sample of scatter plot
38
o Special-to-tariff calls are very succesful because it
is seen that special-to-tariff calls are nearly 10 times more used than inside-network,
outside-network or fixed-line.
39
o We estimate that cheaper calling prices or maybe
free calling in inside-network is appliable in Tariff 3. There is no special-to-tariff
application in this tariff. Also we can say that customers of this tariff are active users of
GSM operator which means that these users are using this GSM operator during whole
day.The clues that support our thesis are the average calling times of outside-network
(nearly 2 times more than other tariffs) and average number of messages(nearly 2 times
more than other tariffs). Tariff 3 is more succesful than the other tariffs; Tariff 3 is able
to make the customers to use the GSM operator actively.
40
usage. In this tariff, GSM operator gives customers the chance of selecting limited
number of numbers(2 or 3), and charges callings less.
41
42
CHAPTER 5
CONCLUSION
42
o Too many tariffs exist in the current condition. It
is imposible to promote all these tariffs to potential customers.
43
REFERENCES
44