A Strategy for Web Development in Electronic Trading: A Case Study

holeknownΑσφάλεια

5 Νοε 2013 (πριν από 3 χρόνια και 5 μήνες)

61 εμφανίσεις

A Strategy for Web

Development in Electronic Trading: A Case Study Page
1

Global Networked Organizations

Twelfth International Bled Electronic Commerce Conference

Bled, Slovenia, June 7
-

9, 1999



A Strategy for Web Development in Electronic Trading:

A Case Study



Dasha Klyachko and Steve Smithson

London School of Economics

H
oughton Street

London WC2A 2AE, UK






Abstract

New electronic commerce technologies, often piloted by start
-
up companies,
represent an opportunity and a threat to existing organisations. This paper discusses a
case study of electronic stock trading and
describes how an existing organisation was
able to grasp the opportunity offered by web technology to develop a web trading
system. The paper focuses on the strategy adopted to reduce the complexity inherent
in the existing organisational and systems envir
onment and hence facilitate the
creative and productive development effort required. The paper offers a stage model
of the development of the system that could be generalised to other electronic
commerce contexts.



1.

Introduction


It is becoming almost a cl
iché to emphasise the impact of the fast changing
technological and business environment on today’s organisations. Nevertheless,
particularly in certain areas, such dramatic changes are a significant and inescapable
feature of the organisational landscape.

Our case study concerns financial trading, an
industry that has changed beyond recognition over recent years (Bakos, 1998).
Deregulation, globalization and the introduction of automated trading systems
transformed the traditional stock markets around the
world during the 1980s and early
1990s. Brokers and large organisations were forced to invest enormous sums of
money in information systems in order to remain competitive in this harsh new
environment.

A Strategy for Web

Development in Electronic Trading: A Case Study Page
2


More recently, the introduction of internet and web
technology (web servers, clients,
internet security tools, etc.) offered an opportunity to create a new web trading
channel. Whereas the electronic trading systems of the late 1980s were relatively
expensive, and difficult to develop and operate so their u
se was restricted to specialist
traders, web trading offers a cheap and simple alternative that is attractive to both
individual investors and larger organisations. Many countries have seen a significant
broadening in share ownership, as this form of savin
g has become increasingly
attractive to a wider range of consumers. Thus the time was right for the development
of an inexpensive and user
-
friendly retail channel for stock trading. Bearing in mind
the nature and culture of the web, it was not surprising t
hat the early web trading
systems were developed by smaller companies, such as E*Trade. Their relatively
quick acceptance by the public threatened a further shake
-
up in the already embattled
financial services industry.


This paper describes how a large US

financial corporation (which we shall refer to as
FinCorp) responded to the challenge of the new web technology and the threat of the
new entrants into the industry. The paper is based on the experiences of one of the
authors who worked on FinCorp’s web
development and witnessed the process at first
hand. For simplicity and clarity, we use Leavitt’s ‘diamond’ (described below) as a
conceptual framework to analyze the events and actions of the organisation. The
‘stage model’ that we offer is based on our r
eflection on the development process and
is not a project management methodology employed by FinCorp.


2.

Conceptual framework


Isakowitz
et al.

(1998) argue that the development of web information systems (WIS)
requires a rather different approach from that
of traditional IS development
methodologies. They attribute this to the tight integration needed with other
information systems, such as databases and transaction processing systems, as well as
the much wider audience addressed by these systems. This techn
ical and business
integration is required both inside the host organisation and with the organisation’s
trading partners; in other words, it needs to fit the interorganizational situation.
Furthermore, WIS are often developed as a result of initiatives and

ideas from the
grass
-
roots of the organisation.


This suggests that the evaluation and analysis of WIS development requires a broader
spectrum than is normally employed. For this reason, we adopted Leavitt’s (1964)
‘diamond’ as a conceptual framework to a
nalyse our case study. This model considers
the interaction between people, processes, technology and structure (Fig. 1). A change
in any of these four elements impacts the other three elements, which should adjust to
absorb the change, and hence benefit t
he organisation. Where one or more elements
does not (or cannot) adjust, this is likely to constrain the change.


Figure 1

A Strategy for Web

Development in Electronic Trading: A Case Study Page
3

Leavitt’s
Diamond
Processes
Technology
People
Structure
Business
Environment
Technological
Environment




Thus, in considering the development of WIS, we should consider not just
technological changes but al
so their interaction with the organisation’s people,
processes and structure. Therefore, there are additional concerns apart from the
technical requirements of building a new channel (e.g. conversion to IP
-
based
networks, installation of new hardware and s
oftware). We need to be concerned with
the training or hiring of people, the processes underlying development and operation,
and any changes needed to the organisational structure (e.g. creation of a new
business unit) within which the development takes pl
ace.


While we would argue that traditional IS development methodologies are
inappropriate for WIS, the underlying notion of dividing the development activity up
into stages is still valid. This compartmentalisation into stages has been shown to be
success
ful for interorganisational IS development (Ramanath
et al.

1998) as well as
fitting the incremental approach, based on learning from pilot developments,
recommended by Bjorn
-
Andersen & Chatfield (1996)


3.

Case study: context, business problem and the strate
gy
adopted


FinCorp is one of the largest discount brokers in the US, with retail customers being
the majority of the client
-
base. FinCorp is perceived to be technologically progressive
among other brokers and financial institutions. The company has a lar
ge central IT
department, aligned with the structure of the business units (e.g. retail, institutional,
etc.). The IT department specialises in system development, integration and
maintenance, with somewhat less emphasis on development. The back
-
end system
s
environment (e.g. trading modules and customer accounts, transaction history,
databases) is primarily mainframe
-
based with occasional UNIX servers. The client
(desktop) software runs on NT. The nation
-
wide network connects the data centre(s)
with the var
ious branches and service centres. The original mainframe remains the
major platform for trading applications, and so the ‘younger’ UNIX/C/C++ mentality
in system development does not prevail. Similarly the IT processes are ritualised and
procedure
-
orient
ed.


By 1995, when FinCorp developed its first web
-
trading site, it was already an
established leader in retail discount brokerage. FinCorp already provided trading
through customer representatives as well as electronic trading via automated phone
A Strategy for Web

Development in Electronic Trading: A Case Study Page
4

systems
and dial
-
up software. However, it had become clear to the company’s senior
management that they needed to respond to the threat to its existing business channels
represented by developments in web technology. The introduction of this technology
(e.g. web
server, client, security, etc.) represented a new business opportunity to
conduct trading on
-
line. This opportunity was available both to existing brokerage
houses (e.g. Charles Schwab, WellsTrade) and the new start
-
up companies that had
been created to op
erate in this new space (e.g. E*Trade, Ameritrade).


To offer a new trading channel an existing company needs either to enhance the
current IS or create a new system. The technological change needed for web trading
was significant enough that existing sys
tems could not simply be altered, a new
information system was required. However, in order to ensure consistency between
the web and other trading channels, the new web trading system had to be integrated
with the existing trading modules.


In competing w
ith the new start
-
up web traders, existing brokers like FinCorp had the
advantage of a large existing client base as well as considerable expertise in stock
trading. On the other hand, they suffered from a complex system environment and
slow bureaucratic o
rganisational structures inherited from the previous trading
channels and other brokerage
-
related systems. Any significant change was likely to
encounter constraints at least, if not outright resistance from the existing
organisational groups. The complexi
ty of trying from the start to integrate the new
web trading with this legacy was likely to have a negative impact on both the old and
the new systems. In this regard, the new start
-
up companies had an advantage, even
though they lacked the customer base.


FinCorp’s senior management accordingly resolved to neutralise any resistance and
facilitate the change process by creating an almost separate, small sub
-
organisation
called Web Trading, within the large organisation, that would be responsible for
develop
ing the new web trading system. Furthermore, it was decided to minimise the
technological legacy of complexity by purposely excluding some components and
initially developing only the most essential modules.



4.

Stages of the development process for the web
trading
system


The authors’ reflections on the development path suggest that it can usefully be split
into five stages. This section briefly outlines each stage, as seen from the perspective
of FinCorp’s web system analysts.


Stage 1 Start
-
Up


In this st
age FinCorp adjusted to the new web technology and began to translate the
old services into the language of the new technology. The business aim was to
establish a presence in the market, in terms of developing a new, but fairly primitive,
web trading syst
em with limited functionality in a short time frame. At this stage web
A Strategy for Web

Development in Electronic Trading: A Case Study Page
5

trading was only an interface to the existing back
-
end systems, a new channel added
to the existing ones (such as telephone and dial
-
up). All channels were equally
dependent on the back
-
end data bases and the trading module.


In terms of organisational structure, FinCorp ‘faked’ a separate start
-
up
-
like, small,
and mobile group (Web Trading), encapsulated in the body of the parent organisation,
and operating with relative freedom from t
he global organisational business risk and
security procedures. To minimise the influence of the existing structure, the new unit
was physically separated from the central IT department by moving to a different
location, close to the electronic trading bus
iness unit. A new executive officer was
hired from outside to lead the unit.


To accelerate the development, further ‘faking’ produced a clean start
-
up systems
environment by artificially redu cing the number of components involved in
integration. While th
e functionality was also reduced because of that, the main goal of
establishing market presence was achieved. IS design and implementation feature
‘home
-
made’ databases replicating
1

the corporate master copies of data and
unsophisticated volume management
and queuing algorithms.


Apart from the initial code development and integration with the system modules,
development tasks included building the infrastructure, security (firewall and
procedures) and IP network for web trading. The company hired new staff

and trained
existing staff with new skills such as UNIX system administration, network security
and HTML, C, and Java programming.


Thus, in terms of Leavitt’s model, the new sub
-
organisation possessed a very different
profile (people, technology, proces
ses and structure), compared to FinCorp itself.


Stage 2 Institutionalisation



Once the new web trading system established a presence in the market, the next target
became the institutionalisation of the new trading channel within the organisation.
There

was a need to integrate it with the information systems that were purposely
omitted, as well as building organisational support structures and incorporating the
‘start
-
up’ unit into the structure of the organisation. The aim was to fully utilise the
suppo
rt mechanisms of FinCorp, to grow the unit within the boundaries of the
organisation, and to reduce the risk

2

of change control or application errors by using
the established procedures.





1

The replication of data bases is the way to reduce t
he complexity of the existing
environment by avoiding the addition of the new data structures to the existing data
bases. The static data storage for the symbol look
-
up functionality is a way to avoid
integration with the security master.


2

By its nature
the electronic trading IS are visible to the ‘external’ world

e.g. customers and analysts) which makes the impact of a failure larger.

Internet has
an ability to magnify mistakes and make then intensely public

. (
Willcocks et al.
1999)

A Strategy for Web

Development in Electronic Trading: A Case Study Page
6


At this stage web trading went through a quantitative change when
it started to
generate a significant level of revenue. The customer base and the processed volumes
increased making the web system visible and, therefore, increasing its operational
risk. The company could not continue to support the initial independence o
f the unit
when it represented so much of the organisation’s business. Other channels did not
allow this new channel to pose such risks to their own survival. To reduce these risks
FinCorp attempted to institutionalise the new service by forcing it to foll
ow the
traditional operational procedures. This resulted in less innovation in system
development and conservatism in the change process. Although the original unit itself
grew, the control of the development was taken away to become company driven.
Respon
sibility and control shifted from being the personal responsibility of a
particular person, while the unit was small, to become a group responsibility at the
level of senior management. Thus, ownership became less clear and was procedurally
driven.


The o
rganisational structure of the web trading unit was also changing. To satisfy the
increasing development and support requests, the Web Trading was growing in size.
The specialised sub
-
units (web trading, web context, web design) started to emerge
within th
e original unit increasing the communication and decision
-
making paths. The
relationships with the main IT infrastructure groups (such as change management and
operational support) were becoming much stronger. While web trading remained the
main business p
artner, the relationships with other business units were also starting to
develop.


IS development at this stage was intensive, including the integration with corporate
back
-
end systems and databases, the development of a better (graphical) user interfac
e
and improved volume management.




Stage 3 Creation of web
-
specific features


Until this stage the functionality of the new system mimicked the functionality
available through the other trading channels. It was only at this stage that the features
speci
fic to the new technology started to emerge. This emergence became possible,
firstly, because the technological possibilities were better understood by then and,
secondly, because the core trading services function was supported by the entire
organisation.



This allowed the development group to focus on the new web
-
specific services
designed and built on the border of the new technology and the old business services.
Examples include the customisation of an account view, quote charts, watch lists, and
a co
mmissions calculator. These were not available to the customers using the
traditional channels. These changes were added independently by various web trading
sub
-
groups. The structure of the business unit, therefore, branched out, reflecting the
split betw
een the core business functionality (trading) and the value
-
added services.
The user interface was also further improved.


Stage 4 Modification of the business and technology


A Strategy for Web

Development in Electronic Trading: A Case Study Page
7


Two interconnected events happened at this stage: the modification of the busi
ness
and the corresponding modification of the web technology/systems. Through the
enlargement of the functionally available context achieved in the previous stage, web
trading went through a qualitative change. Essentially it became ‘double
-
click’
trading
. Often the entire trading process (e.g. information analysis, trading
transaction, payment) happened in the ‘web
-
world’.


The fully electronic and double
-
click nature of web trading allowed increased trading
volume and faster trading. "The North American

experience shows that online
investors buy and sell up to four times more frequently that those using telephone
brokers." (Financial Times 2/1/99, "The Next Big Bang" by James Mackintosh). The
demand for web trading grew faster than the information syste
m providing web
trading. The response time (at the peak hours or for Internet
-
based IPO) became a
problem. The technology that originally created the new space for business
opportunities had to be modified to meet the volume of transactions generated. In
addition to the volume problems, other business requests/modifications (such as
Internet IPO, international trading, and mergers) drove the system development at this
stage.


Arguably, at the time of writing, FinCorp is currently situated within Stage 4.
The
initial web trading site was created in 1995 and by 1998 50% of all stock transactions
were being processed through web. Nevertheless, there is discussion about the future.


Stage 5 Web trading becomes mainstream and is outsourced



Assuming that web

trading retains (or increases) its importance, and there is no
evident reason for change, one may speculate about the future. As such systems
become increasingly common, there may be little commercial advantage in retaining
all aspects of the system in
-
ho
use. One can foresee firstly the outsourcing of its
maintenance and graphic design to specialist suppliers, and eventually, system
development and security control.


5.

Discussion


The results of the case study show that both the web trading system itself a
nd the
development method adopted were highly successful. In order to assess the success
we use the framework of hard, intangible, indirect and strategic benefits (Giaglis &
Paul, 1998). By 1998, 50 per cent of all transactions were being processed online.

In
terms of ‘hard’ benefits, revenue from trading increased while branch costs were
reduced. The most important intangible benefit concerned customer retention. Most of
the customers who wanted to trade online and would probably have been lost to the
comp
etition, stayed with the company. Some customers were lost due to the lower
transaction charges offered by the start
-
up competitors but this was not a serious loss.
FinCorp is sufficiently prestigious that it does not need to be the cheapest trading
servic
e in the market. Indirect benefits included the relative ease and the reduced cost
of building further web trading systems. Internet
-
specific features such as
A Strategy for Web

Development in Electronic Trading: A Case Study Page
8

personalised and customised product offering, could be developed on top of the initial
web tradin
g system. Finally, strategic benefits could be seen in a significant market
share in the virtual business space.


It was generally agreed by the developers involved that the web development method
described above was highly successful. Isolating the initi
al development effort from
the rest of the IT department had produced a flexible, creative and productive
environment that was not constrained by traditional power struggles, bureaucratic
rules or the ‘mainframe culture’. Nevertheless, once web trading had

‘taken off’ and a
significant volume of transactions was being recorded, there was a need to reduce the
technical and business risks by integrating the new trading channel back into the safe
and secure environment of the parent organisation. Although the
initial web trading
system that was developed had to be re
-
worked in parts (e.g. the user interface and
connections to the databases), this was to be expected while the development group
was engaged in a learning process at a time of frequent technologica
l changes. The
fact that these changes were accepted by both the development group and the parent
organisation is a testament to the flexibility inherent in such a development strategy.


The current IS issues faced by the company (such as high volume proce
ssing, slow
response time, security, prompt development of new features, and the maintenance of
redundant servers) are commonplace within electronic commerce systems and were
not a result of the implementation path.


We feel that the development strategy
described above can be generalised to other
sectors facing similar problems of an excessively complex IS environment and
significant market pressure. This strategy of artificially reducing the complexity
engendered by a legacy of technological and organisa
tional bureaucracy by separating
out the new development effort would seem to be relevant to many electronic
commerce developments. In many business sectors, the rapidly changing technology
and complex inter
-
organisational settings threaten to constrain su
ch development
projects. This notion of incremental development is clearly related to the generic
notion of prototyping (
Alavi, 1984
) and the pilot approach recommended by Bjorn
-
Andersen & Chatfield (1996).


This strategy can ease the reluctance of organis
ations to adopt electronic commerce as
the organisational changes are staged over time:

“One of the main reasons that may explain the reluctance of the organisations
to adopt E
-
Commerce on a great scale may be the significant amount of
organisational chan
ge required.” (Giaglis & Paul, 1998)

In this setting, the staged approach suggests fast deliverables and less risk. According
to Hubbard (1998) the risk of project cancellation or failure grows with the length of
the project and can reach 50% failure rate
for projects of 10,000 work/months.


6.

Conclusion


New web technology offered both a business opportunity and a threat to FinCorp, the
subject of our case study. The strategic response of the organisation was to create a
A Strategy for Web

Development in Electronic Trading: A Case Study Page
9

web trading system. The strategy and
management structure adopted was to set up a
separate development group and build the new trading system in an incremental,
staged approach. These stages, comprising start
-
up, institutionalisation, creation of
web specific features, modification of the bus
iness and the technology and (possibly)
outsourcing, involve both systems development and the underlying business. They
permit a much more effective, innovative and flexible approach to the development of
electronic commerce system s.


Further research wil
l concentrate on the changing nature of web trading (increased
personalisation and customisation of product offering, ease and appeal of web trading)
and how this influences IS development. We would also like to explore the effects of
the standardisation a
nd development frameworks on web system development. It
would also be interesting to test the generalisation of this strategy on an example from
another industry.



References


1.

Alavi, M. (1984) An assessment of the prototyping approach to information
syst
ems development.
Communications of the ACM,
Vol.27, No. 6, 556
-
563.


2.

Bakos Y. (1998) The Emerging Role of Electrnic Marketplaces on the Internet.
Communications
, Vol. 41, No. 8.


3.

Bjorn
-
Andersen N. and Chatfield A. (1996) Driving Organisational
Transformati
on through the use of Inter
-
Organisational Systems (IOS).
Electronic
Commerce for Trade Efficiency and Effectiveness, Bled.


4.

Giaglis. G. M. and Paul R.J. (1998) Dynamic Modelling to Assess the Business
Value of Electronic Commerce
. Electrnic Commerce in th
e Information Society,
Bled
.


5.

Hubbard D. (1998) Hurdling Risk.
CIO Magazine, June 1998
.


6.

Isakowitz T., Bieber M. and Vitali F. (1998) Web Information Systems,
Communications Vol. 41, No. 7
.


7.

Leavitt, H.J. (1964) Applied organization change in industry: St
ructural, technical
and human approaches. In: Cooper, W.W., Leavitt, H.J. & Shelly, M.W. (eds.)
New Perspectives in Organizational Research
. Wiley, 55
-
71.


8.

Willcocks L. P. and Lester S. (1999) Information Technology: Transformer or
Sink Hole.
In Willcocks
L.P. and Lester S
.
(eds.)

Beyond The It Productivity
Paradox.
Wiley
.


9.

Ramanath A.M., R.J. Paul, R. Macredie (1998). Understanding Interorganisational
System Development.
ECIS' 98 Proceedings Vol. IV
.