Other Free Encyclopedias » Online Encyclopedia » Encyclopedia - Featured Articles » Contributed Topics from P-T

Technology for E-Marketing - Background, Building a Theoretical Framework Using an Action Research Methodology, Input, Processes, Strategic Output

data customer internet information

Firms have only just begun to fully use the Internet to obtain
customer information in their database marketing processes to enhance
customer relationship management (CRM). This chapter introduces a
framework about how they can do this. Essentially, it argues that the
advent of Internet/database marketing brings solutions to some of the
difficulties in customer relationship management by providing
one-to-one interactivity and customization. For example, the Internet
offers benefits, such as
Page 158 
consumer data collection accuracy and speed, cost savings in collecting
data, greater interaction, and better relationships with customers.
This chapter develops a framework for integrating the Internet and
database marketing to help marketers improve customer relationship
management through rigorous action research.


The growth in database marketing and the emergence of e-commerce
driven by the exponential growth of the Internet requires marketers to
capitalize on the full advantage provided by information technology to
be competitive. The key component of database marketing is its ability
to enhance an organization’s marketing program by identifying customers
that are likely to be more receptive to a specific offering. Indeed,
competent database marketing practice needs to be integrated with other
marketing strategies and practices. The interactive Web environment and
the advent of Internet marketing present an explicit opportunity for
firms to achieve maximum database marketing benefits.

Although the basic database marketing principles are the same, the
integration of the Internet into database marketing process allows
personalized interaction and communication. That is, the Web medium
enables the capture and use of highly personalized information, such as
name, interest, type of car owned, TV programs watched, and so on. This
information, in turn, facilitates one-to-one marketing (Gillenson,
Sherrell & Chen, 1999). For example, armed with customer
transaction data and/or third party lifestyle data from companies like
Claritas and Acxiom, online direct marketers can deliver personalized
interactive promotions, realizing full capabilities of the Web.

A prerequisite for the successful translation of the relationship
marketing paradigm from industrial to consumer markets is accurate
customer information. Improved quality of customer information enables
marketers to target their most valuable prospects more effectively,
tailor their offerings to individual needs, improve customer
satisfaction and retention, and identify opportunities for new products
or services. Therefore, the key focus of e-marketing is customer data
that can be used to inform operational, tactical, and strategic
decision making (Chaffey et al., 2003).

The Internet offers a valuable opportunity to collect information
about a customer (Rowley, 1999). Through the Internet, every customer
contact canPage 159 
be used as
an opportunity to collect data about the customer, which in turn is
incorporated into the marketing database to develop completed customers
profiles. Such information can then be matched to numerous databases
internal or external to the organization, yielding rich permutations of
consumer profiles at a minimal cost. This combined data can then be
used to build customer knowledge which drives marketing strategy to
develop practical long-term relationships with customers. These
relationships, if successful, will lead to further detailed
customer-provided data and/or sales that yield more data and therefore
start the circle over again (Hughes, 1991). The two-way interactive
nature of the Internet mechanisms presents an opportunity for firms to
increase the speed and volume of this circle of customer data

In brief, Internet technology has made it relatively easy to collect
vast amounts of individual customer information (Prabhaker, 2000). Data
quality, entity recognition, synchronization, and integrated databases
enable firms to target content to demographics so precisely that they
can reach markets as narrowly defined as a single customer. This
results in greater levels of customer satisfaction and increased
organizational learning. However, merging the off-line and online
databases raises issues, such as technology compatibility, data quality
and format variances, how to use the sheer volume of data collected,
and consumer privacy concerns. This chapter aims to develop a framework
about how the integration of database and Internet marketing can be
applied, based on rigorous research.

Building a Theoretical Framework Using an Action Research

An action research methodology was adopted to build a framework
about integrating Internet and database marketing. Action research is a
cyclical process methodology that incorporates the processes of
planning, acting, observing, and reflecting on results generated from a
particular project or body of work (Dick, 2000; Zuber-Skerritt &
Perry, 2000). The concept is principally concerned with a group of
people who work together to improve their work processes (Carson et
al., 2001), such as the identification of apposite strategic outputs
from the integration of Internet/database marketing to support

The choice of action research as the methodology was based on two
factors. First, due to the minimal research that has been conducted
regarding the appropriate process for integrating customer information
for e-marketing, the manner through which this may be effectively
completed was unclear. Thus, exploratory research was required, and
action research provides this capability better than many alternatives
(Dick, 2000; Zuber-Skerritt & Perry, 2000). An action research
project within a major multinational entertainment institution was used
to discern the issues involved. The second reason was the flexibility
afforded by action research, which was significant in the
implementation of a research methodology within an evolving information
technology project concerning a problem about which little was known
(Carson et al., 2001).

The action research methodology commences with a group jointly
concerned with a successful conclusion doing fact-finding about a
particular problem in which existing research and understanding is
minimal (Altrichter et al., 2000; Edwards & Bruce, 2000). This
group in the researched multinational entertainment company involved a
project team concerned with the integration of customers from across
various entertainment divisions and used customers’ explicit and
implicit interests to drive a customer-centric e-marketing strategy.
The company’s divisions are vast, including free-to-air and cable
television, print, Internet portals, gambling, sports, show
entertainment, and so on. The project team consisted of
middle-to-senior management of the entertainment company in addition to
the researchers; the joint goal was how to best collect customer data
through the Internet, the integration of this data with on and off-line
databases, and the best e-marketing strategies to implement.

The project team agreed that Internet and database marketing offers
the company a great opportunity to gather market intelligence, monitor
consumer choices, and achieve a closer client relationship through
customers’ revealed preferences in navigational and purchasing
behavior. An initial plan for online data collection, integration with
off-line data, and preliminary marketing strategies to implement was
formulated (Edwards & Bruce, 2000). In particular, a framework
outlining the process through which Internet-derived customer data may
be effectively integrated with other customer data to enable competent
database marketing was discussed and modeled in Figure 1. The framework
has the three usual parts of a system: inputs, processes, and outputs.
That is, inputs of information about customers gathered via the
Internet are processed and converged with a firm’s customer databases
to produce a series of strategic outputs. Consider these three parts in
more detail.


A logical place for the entertainment company to commence collecting
customer information through the Internet was their established
customer databases derived from more traditional customer contact,
billing, and marketing mechanisms. The main customer data categories
that could be collected both online and off-line for competent database
marketing were listed in Table 1 after a brainstorm session.

The project team agreed that for the conduct of effective
e-marketing, the data items in Table 1 need to be collected at a
spatial and/or longitudinal detail. This detail allowed the customer to
be effectively recognized at each transaction, and for a profile of the
customer to be increasingly developed and enhanced over time based on
their response history to online content, expressed preferences and
previous offers (Seybold, 1999). This approach, in combination with
using cross-promotional offers of its other entertainment divisions and
offerings, was designed to enable the entertainment company to
progressively realize the benefits of e-marketing. The development of
this customer profile and associated interest/purchasing patterns were
broadened and expedited by utilizing the following:

  • modeling of customer relationships, scenarios, and propensities
    using investigation models, such as regression and interest profiling
    based on online behavior;

Primary Data Category Example Data Elements
Recognition data Customer unique identifier
Address (home, work, phone, e-mail)
Date of birth
Descriptive data Income
Total investments
Marital status
Children (number, age, gender)
Transaction history Amounts spent
Purchase category
Channel (online)
Direct preference measures Purchase detail (product, vendor attributes)
Provided preferences and permissions
Response history (to types of marketing messages and product offers)
Measure of media used
Questionnaires (online and off-line)
Cookies (determining a specific user or a return user and measuring
time spent and where at site and entry and exit points)
Trigger events Indicators of life events (relocation, college graduation,
child birth, saving deposit for house)
Measures of product information browsing or requests
Cluster product purchase event (for example, home purchase drives need
for home and contents insurance)
Direct client interest (client informs company of interest in
purchasing product/service)
  • data mining and online analytical processing (OLAP) to elicit
    evolving patterns of usage, trends, and customer life-cycle development;
  • determination
    of consumer financial value to the organization through measurement
    against profitability archetypes, such as recency, frequency, and
    monetary purchase values (RFM analysis) and customer tenure, derived
    revenue, and cross-sell opportunities (TRC categorization);
  • visualization of data through graphics to make more apparent
    trends and tendencies; and
  • artificial intelligence, such as neural networks and complex
    adaptive systems theory.

The output from these methods is then integrated back into the data
components outlined in Table 1 to provide an ongoing and evolving
customer profile that is further enhanced by integrating auxiliary
customer interaction and data collection. The integration process is
complicated however by the often unstructured nature of data collected
through the Internet, the diverse nature, quality, and format of
existing company customer data, and the large volumes of data to handle
in an e-marketing context. The details of this integration process are
discussed next.


The information gathered from the Internet was processed through
identification, standardization, de-duplication, and consolidation
procedures, and a unique persistent database key was applied to each
individual customer (Inmon, 2002). Records were stored in an
organizational data warehouse and updated automatically through data
derivation methods, such as Web form logins (for identification) and
cookies (to track interests). Figure 2 shows the process through which
customer data were integrated and then used for e-marketing. The
integration processes in Figure 2 can be summarized into three steps:

Step one: data content identification and understanding. The
first step of data integration is data identification and understanding
of utilized data content. This step involves the extraction of data
feeds in structured format (for example, in XML or comma/pipe separated
values) from the company’s legacy systems and each of the disparate
data sources, such as existing customer data, public data, and
purchased data collected by a third party. This data was then subjected
by the project team to a data quality audit, leading to a full
understanding of the requisite data attributes. A standard data
definition and associated metadata standard was hereafter developed for
the applicable data attributes, and the e-marketing data elements were
prepared for standardization in this format (Corey et. al., 1998;
Dyché, McKnight & Adelman, 2003).

Step two: Data integration and data aggregation. In this
second step, data from the company’s internal systems, as well as
utilized external data was converted into a standard format using the
metadata and conversion standards developed in step one. Following this
conversion, data attribute correction and verification were undertaken,
wherein invalid characters/records were flagged, errors were corrected,
and important contact/recognition attributes, such as address and
e-mail, were verified (Canter, 2002; Todman, 2000).

The next phase, data matching and linking, involved the company
using specialized software and associated processes to identify
duplicate records and logically joining these through a common
persistent identifier or link (English, 1999; Todman, 2000). The
matching utilized by the entertainment company consisted of two

  1. the initial progression, implemented at the original data
    warehouse data load, which was effected as a batch process; and
  2. ongoing updates, or deltas, which were a combination of real-time
    and batch processes.

Following the matching and linking process, de-duplication was
performed and the data consolidated. From here, semantic
transformations, such as address attribute reformating to facilitate
the data load process into the data warehouse was effected, with
particular attention paid by the project team to effect these
transformations in real-time to allow the system to cope with its
continuous rapid data updates. At this point, the consolidated data is
ready for loading into the data warehouse repository.

Step three: The data warehouse. Following completion of steps
one and two, the cleansed de-duplicated data was loaded into the
central data warehouse. The data warehouse is an enterprise data
construct that integrates, stores, and maintains customer data derived
from organizational legacy systems and external sources and acts as the
cornerstone for e-marketing strategies and operations through its focus
on customer related information (Van Dyk, 2002). This data repository
was large, and therefore, the project team decided to create data
marts–specific data views or subsets of the data warehouse optimized to
a particular organizational division’s needs to assist users in
accessing this information.

Strategic Output

The final step involves the implementation of e-marketing
applications to drive strategic output. Customer data collected through
the Internet was used in such application as:

  • reporting and modeling applications to facilitate customer
    understanding and identifying interests, trends, and propensities;
  • campaign
    management, to distinguish customers according to likelihood of
    responding to a specific campaign initiative, disseminate the campaign
    via e-mail or customized Web pages (refer below) and collect and report
    feedback; and
  • real-time analytical/content personalization,
    which was used to predict customer interests based on provided
    information and Web page click stream or navigation behavior and use
    these interests to deliver personalized Web content and e-marketing

The project team agreed that integrated customer data collected from
online and off-line sources needs to be used more strategically than
just as a tool for conducting targeted promotional campaigns. After a
brainstorming session, four e-marketing strategies were identified most
likely to be derived from the integration of Internet and database
marketing to improve customer relationship management and ultimately
implemented. These strategies are:

  1. Prospecting new customers through affiliate or “stealth”
    marketing and networking
    The wide range of entertainment avenues and associated product/service
    offerings provided by the company, in combination with those of
    partners and advertisers, provided a compelling Web site, which lured
    many prospective customers through a combination of off-line brand,
    special offers, and rewards for registering/customer online
    identification and interest specification. For example, the
    entertainment company incorporated Web content with popular television
    shows and magazines also owned by the entertainment company and offered
    holidays, subscription savings, and special interest related purchases
    via this Web content in return for the customer providing personal
    details and effecting purchases. This combination of offering off-line
    brands through a data-driven Internet marketing strategy proved
    successful, with the new Web site growing rapidly to become one of
    Australia’s most popularly visited.
  2. Promoting and advertising pertinent to a customer’s identified

    The strategy of affiliate marketing and rewarding customer
    identification and interest specification allowed the company to
    structure promotion and advertising to be more closely aligned with the
    customer’s real interests. This alignment was evidenced by increased
    network traffic and uptake of offered advertising. An example is the
    entertainment company structuring its advertising to be associated and
    driven by the specific Web pages in which customers were browsing, in
    that specific sports promotions were associated with specific sports
    pages, such as cricket, rugby or golf, holidays with lifestyle travel
    Web pages, and so on.
  3. Cross-marketing: Further to
    advertising according to a customer’s interests, as discussed above,
    the data integration of Internet customer data with legacy/external
    information allowed the entertainment organization to affect
    cross-marketing initiatives. The entertainment company, for example,
    used online consumer data integrated with off-line magazine
    subscription information to drive special promotions for lifestyle
    offers associated with subscribed magazines through e-marketing.
    Further, the defined interests provided by customers via the Internet,
    in combination with the customer data profiling/analysis tools
    discussed previously, allowed for express interests to be extrapolated
    into cross-sales opportunities.
  4. Retaining customers: The
    project team determined that customer retention could be achieved by
    the improved customer understanding and service factors derived from an
    integrated Internet/database marketing solution. Thus, the
    entertainment company would utilize initiatives, such as owning print
    media subscription discounts, related TV and sports special offers, and
    holiday promotions linked with their reality/lifestyle TV shows that
    were available only to members through its e-marketing channel. Again,
    based on customer sign-up, repeat visits, and repeat purchases through
    the Internet media, this strategy was also deemed effective by the
    project team.
Teisserenc de Bort, Léon Philippe [next] [back] Technology - Raw Stock, Studio Machinery, The Laboratory, Film Exchanges, Theaters

User Comments

Your email address will be altered so spam harvesting bots can't read it easily.
Hide my email completely instead?

Cancel or