METHODOLOGY ON ELECTRONIC CUSTOMER RELATIONSHIP
MANAGEMENT
This chapter focuses on the research method that was used for this study. It addresses the
research design, how the data was collected and how the data analysis was conducted.
The cross sectional survey design gathered information on electronic customer
relationship management from the customer care consultants and managers of
commercial banks in Nairobi, Kenya and its environs. Stratified sampling was used to
select the subjects that represented the population. The target population was employees
of commercial banks of Kenya in Nairobi and its environs. A structured questionnaire
was used to collect data on electronic customer relationship management on competitive
advantage. Descriptive statistics was applied to give central tendency measures such as
mean scores and measures of dispersion such as standard deviation and variance. The
analyzed quantitative data is presented in tables.
3.1 Research Design
This research study adopted a cross sectional survey approach on the influence of
electronic customer relationship management on competitive advantage. The research
was based on commercial banks in Kenya. The cross sectional design best illustrates the
relationship and analyses of factors that support the subject matter being researched on.
According to Kothari (2009), a cross sectional study is interested with discovery of the
what, where and how of an issue. Cross sectional research design is selected because it
empowers the researcher to generalize the findings to a larger population.
According to Mugenda and Mugenda (2008) it is paramount to use data where elements
are observed in either natural environment without influencing them. It is best utilized
when collecting information about people’s perceptions. It is an effective method of
gathering information needed to describe the perceptions and views of bank employees in
customer care and information technology department on the influence of electronic
customer relationship management on competitive advantage.
A cross sectional research design seeks to give a perfect profile of persons, events or
situations by evaluation and analysis of ideas. Commercial banks within Nairobi were
selected for the study because of their extensive branch network and their commitment to
drive change through the adoption of technological electronic platform. Descriptive
survey design allows for sampling thus reduce time used and cost of conducting a study
targeting a huge population hence generalization of the finding (Mugenda & Mugenda,
2008).
3.2 Target Population
Refer to the actual population in research from which information is desired. According
to Ngechu (2009), a population is holistic set of people, services, elements, events or
group of subjects that are being investigated. Population studies are more holistic because
all elements have a fair chance to be included in the end sample that is chosen according
to Mugenda and Mugenda (2008). The target populations of the study were employees
working in commercial banks within Nairobi and its environs.
The study focused on the different departments and particularly on employees who
directly deal with customers since they are the ones conversant with the customers' needs
and manage the customers’ relationships. Mugenda and Mugenda, (2008), explains that
the target population should have some identifiable characteristics, to which the
researcher intends to generalize the results of the study. The definition assumes that the
population sample is not homogeneous. The researcher investigated a sample of
employees drawn from the population of 100 management and general staff working in
commercial banks in Kenya. The population sample can be summarized in table below.
Table 3.1: Categorization of Commercial Banks
Category Percentage Sample size
Large banks (Tier one) 65.98 66
Medium banks (Tier two) 26.10 26
Small banks (Tier three) 7.92 8
Total 100 100
Source: CBK Bank Supervision Annual Report (2017)
3.3 Sampling Design
According to Harper (1991), a sampling frame is the source material or device whereby a
sample is drawn from. A sampling frame is a list of all respondents within a population
who can be sampled, and may include households, individuals or institutions. As defined
by Daniel (1992), a sampling frame is a collection of elements from which a sample is
drawn. Sampling frame provides ways for selecting of particular members of the target
population that are to be interviewed in the survey. The sampling frame of this study was
staff working at commercial banks in Nairobi City County and its environs.
In this study stratified sampling was used to identify the subjects that represented the
population. The sampling method is all inclusive of all members of the population
involved in the study. A random sample is preferred because it is free from bias and
therefore each unit has a possibility to be included in the sample.
This research is a cross sectional study and therefore the researcher studied all
commercial banks in Kenya from the possible 100 target population. The staff and
management are deemed suitable for the study as they have better knowledge and
awareness on the issue at stake and provide specific information from a management,
customer and staff perspective (Kothari, 2008). This study used cross sectional study
formula to calculate the sample as shown below:
n = Z2 *p*(1-p)
d2
n = 1.962*0.93*0.07
0.052
Z = 1.96 (at 95% confidence level)
d= 5% (level of precision)
p= 93% (level of e-CRM usage)
Sample size (n) = 100 respondents
Stratified sampling was used to select the respondents. The strata included the three tiers
in Table 3.1 and within each tier according to the CBK Bank Supervision Annual Report
(2017) in appendix III.
3.4 Data Collection
Daniel (1992) believes that questionnaires are effective because they are efficient in
terms of time utilization, energy consumption and economic value for money. Use of
questions allows for generalization, simplicity and accuracy. The major disadvantage of
the instrument for data collection include; lack of interest, different level of knowledge
gaps, biases, misinterpretation, fatigue, boredom, incomplete questionnaire form are
some of the negative responses affecting use of questionnaires according to Uma. In
order to overcome the problems identified above, the research used both structured and
semi structured questionnaires.
The closed ended questionnaire elements were included in order to increase accuracy to
the questionnaire objectives, and open-ended questionnaires were designed to enhance
inclusivity of the responses during the interviews. According to (Bailey et al., 2008),
questionnaires make generalization of responses easy and provide efficient way of
collecting responses from a large sample prior to the quantitative analysis. Accordingly,
in applying cross sectional design, secondary and primary data will be important for
purposes of comparison.
The researcher used email and printed questionnaire to administer the research instrument
individually to a sample of 100 staff working at commercial banks in Nairobi and its
environs. The customer care consultants and managers provided key information on
electronic customer relationship management. The researcher exercised care and control
to ensure all questionnaires were issued to the respondents and to achieve this, the
researcher maintained a register of questionnaires which were administered and those
which were returned.
3.5 Data Analysis
Quantitative data collected using questionnaires was analyzed by the use of
descriptive statistics using Statistical Package for Social Sciences (SPSS) version
20.0. Descriptive statistics was applied to give central tendency measures such as
mean scores and measures of dispersion such as standard deviation and variance.
Descriptive statistics provided categorizations of valid and replicable findings from
data collected to answer objectives of the research study according to Kirk and
Miller (2009).
The study used level of significance to analyze the degree of association between the
variables. Quantitative analysis generated quantitative reports through percentages,
tabulations and central tendency measures. Quantitative data was presented using
statistical techniques such as frequency counts and percentages to make inferences.
Inferential statistics which included correlation and regression analysis were used.
