Systems Analysis and Design pdf

As we shall see in Chapter 2,structured methods, particularly SSADM, provide a very good discipline inthis regard, as they tend to conduct requirements analysis and requirementsspecificat

Systems Analysis and Design

strongest educational materials in systems analysis,

bringing cutting-edge thinking and best

learning practice to a global market.

Under a range of well-known imprints, including

Financial Times/Prentice Hall, we craft high quality print and electronic publications which help readers to understand and apply their content, whether studying or at work.

To find out more about the complete range of our

publishing, please visit us on the World Wide Web at: www.pearsoned.com

Systems Analysis

and Design

Donald Yeates and Tony Wakefield Second edition

Edinburgh Gate

Harlow

Essex CM20 2JE

England

and Associated Companies throughout the world

Visit us on the World Wide Web at:

www.pearsoneduc.com

First published 1994

Second edition published 2004

© Pearson Education Limited 1994, 2004

The rights of Donald Yeates and Tony Wakefield to be identified as

authors of this work have been asserted by them in accordance

with the Copyright, Designs and Patents Act 1988.

All rights reserved No part of this publication may be reproduced, stored

in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without either the prior written permission of the publisher or a licence permitting restricted copying

in the united Kingdom issued by the Copyright Licensing Agency Ltd,

90 Tottenham Court Road, London W1T 4LP.

All trademarks used herein are the property of their respective owners.

The use of any trademark in this text does not vest in the author or publisher any trademark ownership rights in such trademarks, nor does the use of such trademarks imply any affiliation with or endorsement of this book by such owners ISBN 0273 65536 1

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

10 9 8 7 6 5 4 3 2 1

08 07 06 05 04

Typeset in 10/12 1 / 2 pt Palatino by 35

Printed and bound by Bell & Bain Limited, Glasgow

The publisher’s policy is to use paper manufactured from sustainable forests.

2 Approaches to Analysis and Design 34

2.11.1 Direct Systems Development Method (DSDM) 56

5.2.1 Before Analysis and Design 108

5.4.1 Estimating for Analysis and Design Work 120

Contents vii

9.5.1 The Logical Data Structuring Technique 188

11 Systems Analysis: Modelling Systems Behaviour 217

11.2 Creating a Logical Model of Current Processing 218

11.10 Modelling Enquiries 230

Contents ix

17.3 The Bottom-up View: Third Normal Form Analysis 338

17.6 The Data Dictionary 347

17.6.1 Advanced Features of a Data Dictionary 348

20.3.2 Overheads that Adversely Affect System Performance 38920.4 Investigating the Chosen Hardware/Software Platform 390

Contents xi

20.5 Moving from Logical to Physical Data Design 393

23.3 Change Management 437

This is a substantially new and different edition of Systems Analysis and sign We hope that the new materials bring it up to date and that we’vekept the best of the old It has been written for people studying systemsanalysis and design or who are already working in systems teams Peoplewho are involved in the development of new systems or the development ofnew systems analysts, or both, have written it

De-It is intended to be a practical book, easy to read and easy to use De-It followsthe same structure as the previous edition and has new and updatedmaterial in it There are more case studies and exercises too and we’ve cross-referenced the topics to the ISEB syllabus in systems analysis and design

There is also a web site for the book for the first time We’re excited aboutthe opportunities that this offers We want to make it interesting and useful

to use You’ll find more information about Denton Motor Holdings andsome answer pointers to some of the case study problems and exercises.There are also some end-of-chapter quizzes and the answers to them Youcould use these to check your understanding as you go through the book

For more information, log on to www.booksites.net/yeates.

We hope that you enjoy the book It can’t tell you everything you need toknow about analysis and design; no book can do that There is no substitutefor experience

Finally our thanks to James Cadle, John Koenigsberger, Steve Copson andMurdoch Mactaggart for reviewing some of the material for us, and to thepanellists who helped with the final chapter – Debbie Paul, Frank Jones,Jean-Noel Ezingeard, Nigel Underwood and Richard Bevan Also, we wouldlike to thank Amanda Thompson and Tim Parker at Pearson Education for

keeping our noses to the grindstone in the nicest possible way as well asLionel Browne and David Hemsley the project editors.

We are grateful to the following for permission to reproduce copyright material:

Figure 2.7 ‘Custom Order Workflow’, published by Actionworks® – Business Process Management for People Processes; Fig 4.2 from The Team Handbook by Peter Scholtes, published by Oriel Incorporated; Fig 6.5 from STARTS Guide

1987, supported by the Department of Trade and Industry; Fig 8.10 ‘NCC clerical document specification’, published by National Computing Centre.

In some instances we have been unable to trace the owners of copyright material,and we would appreciate any information that would enable us to do so

1 The Context for Analysis and Design

The book is organised in a project-chronological sequence This means that itbegins with analysis and ends with implementation There are some chapters,however, that don’t fit neatly into this sequence, so they have been put in at thebeginning – like this one – or right at the end – like Change Management Theydeal with the environment within which analysis and design takes place and areconcerned with business, people, management and quality There’s also a basicassumption running through the book It is that analysts and designers work forcustomers We believe that the word ‘customer’ is very important Somebodypays for what analysts and designers deliver New systems have to be justified

by the benefits that they deliver It is easy to use terms such as ‘the users’ and

‘user management’ – they’re used in this book – and forget that they are stitutes for ‘the customer’ All of the contributors to this book have workedfor companies whose very existence depended on their ability to build anddeliver new computer-based systems and where customer focus came first,last and everywhere in between This isn’t to say that only those analysts anddesigners who work for service companies have a customer focus Indeed, thechapter on Building Better Systems extends the scope of the word ‘customer’well beyond its everyday usage and beyond the meaning here in this chapter

sub-We do, however, want you to hold on to the important concept of ‘customer’ as

you read this book There are three other important points that need to bementioned here.

• Systems analysis and design involves people Certainly it involves technology,often technology that we don’t really understand and which we rely on otherpeople to manage for us, but the best-designed systems in the world succeedbecause the people who use them can do their jobs better This is because theirinformation systems enable them to achieve goals they wouldn’t otherwisereach There are just too many people on trains and planes using laptopmachines to believe that we can ever in the future manage without computers.The importance of this ‘people aspect’ is emphasised later where we considerchange management

• Systems analysts and designers change the world This is a bold statement,and in one sense everyone changes the world to some extent just by their veryexistence The role of the systems analyst, however we may describe it indetail, is to be a change agent Unless we are interested in changing the wayorganisations work we have no need for systems analysts Unless you areinterested in changing the way organisations work, you probably have noneed to read this book

• Information systems are expensive to develop and maintain, so it is clear thatbusinesses do not commission them just for the fun of it The need for aninformation system must grow out of some perceived business requirement,and the justification for it must be expressed in business terms Although this

is well enough understood in theory, it is surprising how often IS projects

do start without clear links back to business plans and strategies Systemdevelopers often take the sketchiest of briefs and start to develop somethingthey think meets the need – and are then unpleasantly surprised when theuser or sponsor of the system refuses to accept it because it does not properlymeet their specific business requirements

1.2 Business Analysis

1.2.1 Levels of Understanding

A proper analysis of system requirements requires that those carrying out the vestigation have a balance of skills and knowledge This includes a good generalunderstanding of the operation of business and of the factors that affect theviability of all businesses This involves a broad-based knowledge of finance,accounting, marketing, production and distribution, ideally gained in a variety ofbusiness environments There will also be a need for a more specific grasp of theimportant features of the particular business being studied So, for example, ananalyst working for a high street chain of shops should have knowledge of theretail sector and be able to discuss retailing matters knowledgeably with thecustomer Finally of course a broad knowledge of the possibilities and limitations

in-of information technology is also required

This balance of skills is needed for three reasons First, so that the analystsunderstand what the users of the proposed system are saying to them andcan recognise the implications of the requirements they are capturing Also,because users will not expect analysts to adopt a completely passive role;they will want them to challenge assumptions and interject ideas to stimulatethem in thinking out their requirements Without an understanding of the busi-ness of the organisation, or of how other organisations have tackled similarissues, it will not be possible to play this catalytic role Finally, the users willgenerally expect the analysts to devise and propose technical solutions to theirbusiness problems.

Systems analysts, however, are often generalists and, particularly in servicecompanies, are quite likely to be sent to work with a travel agent on one project,

a bank on the next and a manufacturing company on the third Although thiscan facilitate cross-fertilisation of ideas between different sorts of business, thedisadvantage is, of course, that they sometimes lack the specialist knowledge todeal with users on an equal basis So if this is the case, what can be done to bringrelevant experience to bear on a particular assignment?

First, and most obviously, the project manager should try to find analysts withprevious experience of the business to be studied, or something similar Someonewho has worked on a distribution application, for instance, may have achieved areasonable understanding of areas such as stock-keeping and just-in-time deliv-ery systems, which can be brought into play on a retail assignment If, however,analysts with relevant backgrounds are not available, additional support willhave to be provided by arranging some training for the analysts in the areas

of interest in the form of public courses or self-teach packages, or by hiring anexpert to coach them in the new business areas Alternatively, suitable back-ground reading can be provided Another approach is to provide consultancysupport, which can be used either to lead the investigation with the customer

or to provide background advice and guidance for the analysis team

A particular challenge for analysts, and one that seems more difficult for thosefrom a technical background such as programming, is to keep the business re-quirements as the focus instead of getting hooked up on technological solutions

It is very easy, for example, to take a messy and cumbersome manual system andproduce instead a messy and cumbersome computer system Analysis should atall times be tightly focused on the business objectives that the proposed system

is supposed to fulfil, and only when these are thoroughly clear should theidentification of technical solutions be attempted As we shall see in Chapter 2,structured methods, particularly SSADM, provide a very good discipline inthis regard, as they tend to conduct requirements analysis and requirementsspecification at a wholly logical level and deliberately exclude the consideration

of technical issues until the business requirements have been settled

1.2.2 Linkage of IS to Business Objectives

There are many reasons why businesses should want to develop information tems, but some of the most common objectives are:

sys-Business Analysis 3

• To reduce manpower costs The introduction of computer-based systems has

often enabled work to be done by fewer staff or, more likely nowadays, haspermitted new tasks to be undertaken without increasing staffing levels.The automation of many banking functions, such as cheque clearance, fallsinto this category

• To improve customer service Computer systems can often allow organisations

to serve customers more quickly or to provide them with additional services.Supermarket point-of-sale systems producing itemised bills provide an

illustration of this

• To improve management information Management decisions can only be as

good as the information on which they are based, so many computer systemshave been designed to produce more, or more accurate, or more timely

information With modern database query facilities it is even possible toprovide systems that do not require the data retrieval requirements to bedefined in advance, thereby enabling managers to institute new types ofenquiry when changing business conditions demand new or different

information

• To secure or defend competitive advantage This is becoming a major justification

for spending on information systems, and is examined in more detail later inthis chapter

Ideally, the analyst should work from a hierarchy of objectives, each one posing achallenge to, and imposing constraints upon, those at a lower level The lower-

level objectives are sometimes referred to as critical success factors: that is, they are

things that must be achieved if the top-level objective is to be met The criticalsuccess factors will become more detailed and tightly focused as one worksdown the hierarchy, and perhaps this may be best illustrated by an example.Let us consider a motor-car manufacturing company It currently has, say,10% of the market for its products, and the board has defined a five-year mission

of raising that proportion to 20% But this is a very broad target and, to achieve

it, the organisation will have to define a set of more tangible objectives thatwill lead to its being met – in other words, the critical success factors It may befelt that one key to increasing market share is to offer a more frequent choice ofnew models So, a lower-level objective for the design team may be to reduce thetime to develop a new model from, say, five years to two And the productiondepartment will have to be able to switch over assembly lines in less time, say areduction from six months to three

Coming down a stage further still, the designers will want to introducetechnology that can produce manufacturing instructions, documentation to sup-port the issuing of tenders to subcontractors, component listings and setupinstructions for the assembly lines from their drawings Now, we can derivesome very focused critical success factors for our information system based uponthese detailed requirements It can be seen, then, that business objectives andcritical success factors ‘cascade’ from one level to another, and the whole setshould form a pyramid supporting the overall aims of the business

1.3 Constraints

The range of potential solutions that may be proposed by the analyst will belimited by constraints imposed by the user and by the nature of the user’sbusiness The analyst should ideally undertand these constraints before analysisbegins but certainly as it progresses, and must keep them firmly in mind as theideas for the proposed system emerge

1.3.1 The User’s Organisation

The first thing for the analyst to consider is the structure of the user’s sation It may, for example, operate in a very centralised fashion with nearly alldecisions being made at head office If this is the case, then clearly informationsystems must reflect this pattern and be designed so that relevant informationcan be processed rapidly and presented at the heart of the business If, on theother hand, the organisation allows considerable autonomy to managers in sub-sidiary parts of the business, then the systems must be designed to provide thesemanagers with what they need to run the business effectively In other words thesystems must reflect the structure of decision-making in the business

organi-It is important to remember, however, that organisations are not static andtend to oscillate between centralisation and devolution as circumstances, thebusiness and intellectual climate and the most recent theories of managementgurus dictate Nowadays, the difficulty of altering information systems can prove

a major obstacle to reorganisation Some might think this to be a good thing, but

it is important not to make systems so inflexible that they actually prevent the ganisation from being operated in the way the management decides is necessary

or-1.3.2 Working Practices

It is easiest to introduce a new information system if it leaves existing workingpractices largely undisturbed Conversely, systems that require a lot of changemay prove very difficult to implement

However, the analyst must not allow a fear of the difficulty of implementation

to prevent the best solution – best for the business as a whole, that is – beingadvocated The most radical changes will often provide the greatest gains, and

if that is so then the problems of implementation must be faced and overcome,and the chapter on change management gives some ideas that help in this dif-ficult process Alternatively, the management of the organisation may want touse the introduction of new technology as a catalyst for change, to shake up asleepy, backward-looking department for instance

The key point is that the analyst must make some assessment of the climateprevailing in the organisation Will its management wholeheartedly push forchange, give it lukewarm support, or just run away from its implications?Finally, at a more prosaic level, some working practices that may appear over-elaborate and cumbersome will turn out to have evolved for good businessreasons – such as the maintenance of safety standards on a railway for example

Constraints 5

The analyst must make very sure that these business reasons are not ignored inproposing new and more streamlined systems.

1.3.3 Financial Control Procedures

Various aspects of the way an organisation manages its finances can have animpact on IT developments The first is the concept of capital versus revenueexpenditure Most IT developments involve capital expenditure in that they arefunded as one-off projects rather than out of continuing expenses But the payofffor a capital project may be a reduced continuing revenue cost somewhere down-stream Depending on the organisation’s rules for the ‘payback’ on capital projects,

a capital project cost of, say, £3 million may not be justifiable even if, over a year system life, it could produce revenue savings of ‘only’ £1 million per annum.Also, organisations may have only limited funds earmarked for capitalprojects in a given year but have reasonably generous revenue budgets forongoing work In these circumstances it may be sensible to propose a limitedinitial capital expenditure for a core system, with enhancements and additionsbeing made gradually as funds permit

seven-The analyst needs to consider who actually holds the purse strings for a ticular development and what it is that will convince that person of the worth ofthe proposed development Let us suppose, by way of example, that we are toexamine the requirements for a new payroll system in an organisation The pay-master could be the payroll manager, who wants a fully comprehensive systemthat will enable him or her to offer new services and perhaps even take over thefunctions of the corporate personnel system, or it could be the IT director, who isdeveloping a strategy of packaged systems running on ‘open’ architectures, orthe finance director, who wants a system that will cut the number of staff, andhence the costs, in the payroll department The objectives of each of these man-agers are rather different, and if the analyst is to get a solution adopted, it must

par-be geared to the needs of the person, or people, who will approve and pay for it

1.3.4 Security and Privacy

The analyst needs to determine fairly early on which sort of security conditionswill be required for the proposed information system These could include:

• ordinary commercial confidentiality, where the main aim is to ensure thatsensitive commercial information such as, for instance, the production costbreakdown of products cannot be stolen by the competition;

• more sensitive systems, such as the Police National Computer, where specialconsiderations apply to the holding of, and access to, data;

• very secure systems, such as those that support the armed services and

government agencies

Clearly, the need for rigorous security control could impose major constraintsand development costs on the project

1.3.5 Legal Considerations

It is becoming increasingly the case that the users of information systems areliable for the consequences of things done, or put in train, by those systems

If there are such liabilities, the analyst must examine them and allow for them

in the proposed system Safety-critical systems are the most obvious example,and if one were examining the requirements for, say, an air traffic controlsystem, safety considerations would constitute one of the main constraints ofthe proposed solution Other systems have also fallen foul of the law recently,and it cannot be too long before, for example, a credit-scoring agency is found

to be liable for the consequences of wrongly deciding that someone is a badrisk

So far, too, there has been little legal exploration of the subject of tial losses arising from the use of information systems, and developers have beenable to hide behind contract clauses that limit their liability to the cost of thesystems’s development or some arbitrary figure There must be some possibilitythat this will change in the future, so the analyst, in assessing the risks fromsome proposed solution, ought at least to think about what might be the con-sequential losses resulting in a system failure

consequen-The law does already have something to say on the subject of storing mation about individuals, in the form of data protection legislation This has twoaspects that particularly concern us These are ensuring that the data is held onlyfor defined and declared purposes, and enabling those with a statutory right toinspect information held about them to do so So, if the proposed informationsystem may hold information on individuals, the analyst needs to ensure that therequirements of the legislation can be met

infor-1.3.6 Audit Requirements

An organisation’s internal and external auditors will want access to systems toensure that they are working properly and that the financial information theyproduce can be relied upon They may also require that certain self-checkingmechanisms and authorisation procedures be incorporated into systems In sometypes of system – those supporting pension funds or banks are obvious examples– the need to check and extract audited information forms a large part ofthe requirement itself It is very much better, not to say easier, if these auditrequirements can be taken into account at the specification and design stagerather than added after the system is complete, so the analyst must talk to all therelevant authorities and find out their requirements alongside those of the moreobvious users of the system

1.3.7 Fallback and Recovery

Most information systems have some sort of requirement for fallback and ery These requirements could include the ability to ‘roll back’ the system tosome point before failure and then to come forward progressively to bring theinformation up to date, or some back-up means of capturing data while the main

recov-Constraints 7

system is off-line Standby systems that normally perform less urgent tasks cantake over from ‘critical’ systems, and, if necessary, full system duplication oreven triplication may be provided for critical real-time or command-and-controlsystems Provision of back-up, though necessary, is expensive, and so the case forthe arrangements provided must be examined in strict business terms and theeffects on the business of system failure assessed ‘What would be the costsinvolved?’ and ‘How long could the business go without the system?’ are two

of the more important questions to be answered Recent terrorist attacks havedemonstrated the importance of back-up systems and the speed with which theycan be brought on-line after a disaster

Sometimes the analysis of these consequences can produce truly frighteningresults In one case there was an investigation of a system that supported a majorundertaking and which had distributed data-capture and centralised control

It was found that, if the central processors went out of action for more than twodays, the backlog of data in the distributed processors would be such that thesystem could never catch up, whereas the failure of one of the distributedmachines would not become very serious for several weeks

So, the analyst must carry out a comprehensive risk analysis in this area, usingoutside expertise in support if necessary, and must keep the resultant constraints

in mind at all times

1.4 Using IT for Competitive Advantage

In the early days of information systems, their justification seemed forward enough For the most part the systems were ‘number-crunchers’ thatcould carry out routine repetitive tasks, such as the calculation of payrolls, muchmore quickly and cheaply than an army of clerks The payoff was thus clearly instaff savings plus perhaps some additions in the form of better or more timelyinformation for management

straight-There are few, if any, of these first-time applications available now, and manyadministrative systems are now into their third or fourth incarnations Justifi-cation for the new developments has generally been that:

• The old ones are incompatible with newer technological platforms, resultingperhaps from a switch to ‘open’ architectures

• They have become impossible to maintain because of the poor documentation

or configuration management of many early systems

• They need skills and resources to maintain them that are no longer available

or are prohibitively expensive

None of these however are business reasons in the sense that they support somekey business objective; rather, they are technical reasons justified only in terms ofthe inherent nature of IT itself

As expenditure on IT has risen, so managements have become increasinglykeen to ensure that the money spent contributes in some tangible way to the

achievement of business objectives At the same time, some more enlightenedboards, and some IT directors with a wider interest than in the technology itself,have become interested in the idea of ‘IT for competitive advantage’ The concept

is simple enough If IT can give your company some unique offering, or tribute to providing some unique offering, then it will give you a competitiveedge over your competitors and hence contribute directly to increased sales andprofits Two examples may serve to illustrate the idea, which is discussed inmore detail later in this section

con-Example 1

One of the continuing headaches for retailers is the level of inventory, or stock,they keep in their stores If it is too high, excess funds are tied up in it, profit mar-gins are depressed by it, and it occupies floorspace that could be more usefullyemployed to display and sell goods If inventory is too low, they risk ‘stockouts’,and customers cannot buy what is not on the shelves In high-volume operationssuch as supermarkets, if customers cannot buy what they want immediately,they will go elsewhere, and that particular sale is gone for ever

How can information systems help? A system is implemented that constantlymonitors sales at the checkouts and signals a distribution centre when stocks ofitems fall below predetermined levels Replacement stock is loaded onto trucksand, provided the operation has been set up correctly, arrives at the store just asthe last item is sold The stock is unloaded straight onto the display shelves and,without having excessive local storage space, the store is able to avoid customersgoing away empty-handed Additionally, the cost savings can be passed on tocustomers as lower prices, thus attracting more business and improving themarket share and profitability of the chain

Example 2

In the late 1970s, many airlines were interested in seeing how they could tietravel agents and customers to their services and, by the same token, excludetheir competitors Several of them formed consortia with the idea of developingpowerful booking systems that they could provide to travel agents These wouldmake it very much quicker and simpler for the agents to deal with the participat-ing airlines than with their competitors It is interesting to note that, in the majorshake-up of the international airline business that occurred in the 1980s and1990s, it was the airlines that invested most in this technology that moved todominant positions in the marketplace

In a competitive environment organisations grow and prosper through creasing their competitive – or strategic – advantage over their rivals The best-known analysis of this competitive situation is Michael Porter’s five forcesmodel, which shows how competitive forces impact on an organisation We canuse this model to see how IS developments can help to improve the competit-ive position First, though, because we are dealing with information systems,

in-we need to understand the difference betin-ween data and information, and howinformation is created and valued

Using IT for Competitive Advantage 9

Data is the raw material, the facts and observations of business life that bythemselves have little value or useful purpose Knowing the number of beef-burgers sold today doesn’t really help us to manage the burger shop unless weknow how many were sold on the same day last week, the resources used inmaking them, the price at which they were sold, and so on By contrast, infor-mation is data that has been processed in some way to make it useful, valuableand meaningful and a possible basis for taking decisions In our burger shop wemight for example have processed raw sales data to show sales by each hour ofthe day to help us to determine staffing levels.

So, to get information from data we are:

• Processing or transforming it using a defined process Multiply the number ofburgers sold each day by their price to get the revenue for the day

• Putting it into a context that gives it meaning Sales by each hour of the dayshow the busiest times and predict when most staff will be needed

• Creating it for a purpose We may be trying to decide whether to changeemployee shift patterns

• Using it to help make better decisions through reducing uncertainty and soeliminate the amount of inefficient trial and error

It’s apparent then that information has value, but how is its value assessed?Intuitively managers often say ‘No, don’t do that It’s just not worth the effort

to get that level of detail’, or they ask themselves the question ‘What would I

do with that information if I had it?’ The problem comes when they say ‘What I’dreally like to know is Because then I could and that would mean ’, sothere is a value in having the new information, and although it is difficult toquantify, it is an economic entity and it does have a value The benefits that resultfrom using information might be clearly visible – tangible benefits – such as costsavings Or they might be intangible – soft benefits such as business expansion.Often information takes on value because it generates tangible or intangiblebenefits Cost–benefit analysis attempts to identify the expected payoff from

an information system by quantifying the cost of producing the information andthe value of the information when it’s produced

However the cost benefit works out, information users will be concerned withits quality Let’s look at the characteristics of good quality information:

• Relevant Is the information relevant, bearing in mind how I want to use it?

When I’m presented with it, is extra processing needed before it gives mewhat I need?

• Accurate Wrong or misleading information leads to wrong decisions being

made The information needs to be accurate enough for the purpose; smalldepartments budget items of cost to the nearest £100, larger departments tothe nearest £1000

• Enough Decisions are usually taken in an atmosphere of uncertainty Getting

more information isn’t always the answer either as it may be neither timelynor relevant

• In time Information has much less value if it arrives late! It needs to arrive in

sufficient time for thoughtful decisions to be made

• Clear Is the information presented in a form that makes it easy to understand

and use? This sounds simple: we just ask the user how he or she would like itpresented, and as long as there’s one user with unchanging preferences andneeds then we can deliver it Multiple users with different needs and changingpreferences – which is most likely to be the case – cause more difficulty

A useful way of summarising and remembering the characteristics of good mation is shown in Figure 1.1

infor-Using IT for Competitive Advantage 11

Fig 1.2 The competitive environment

Fig 1.1 The dimensions of good quality information

What part then does good-quality information play in giving organisationscompetitive advantage, whether it means more profit, greater market share, get-ting higher up the government league table, or meeting targets set by an industryregulator? As was said earlier, for organisations to grow and prosper they seek

to gain competitive advantage, and they use information systems to developand sustain this advantage Here’s how Porter’s five forces model works In

the centre is the competitive environment, where there is rivalry among the

existing competitors (Figure 1.2)

This rivalry can be intense or mild, the result of a growing or contractingmarket, where one competitor’s products are highly differentiated from those

of others, or where there are many or few competitors All organisations in the

market are subject to the threat of new entrants (Figure 1.3).

Fig 1.4 The threat of new products or services

The existing competitors may ‘share out’ the market in safety if the threat ofnew entrants is low because the barriers to entry are high In an extreme case itmay be that the capital costs of starting up are very high or it may be necessary toget a licence to operate in the market If existing customers are very loyal to exist-

ing brands, then entry costs may also be high There may, however, be substitute products or services that customers could use instead of the ones currently sup-

plied by the existing organisations If high prices are being paid to the existingsuppliers, who make high profits, then customers may be willing to change Thesubstitute product may also have advantages of its own – real or perceived – thatexisting products don’t have (Figure 1.4)

Fig 1.3 The threat of new entrants

Using IT for Competitive Advantage 13

So now we have three of the five forces: the competitive rivalry, the threat ofnew entrants, and the threat from substitute products or services The remaining

two are the bargaining power of suppliers and the bargaining power of customers.

If there are many suppliers then organisations have more choice and moreopportunity to negotiate good supply arrangements Conversely, if supply isconcentrated with a few suppliers then the greater power they have There mayalso be the costs of switching to a new supplier or being tied to a specific supplierbecause their product or service has special qualities For customers, theirbargaining power depends on the strength of the brand being bought, productquality, the buyers’ concentration versus the sellers’ fragmentation and so on.The final model therefore looks as shown in Figure 1.5

This model of industry competitiveness and the forces in it allows generalquestions to be asked about the role that IS can play So we ask whether IS can

be used, for example, to

• defend a market position against new entrants;

• reduce the power of suppliers;

• reduce the bargaining power of buyers;

• generate greater market share by winning customers from rivals

There are reckoned to be three basic strategies that organisations can follow: they

can compete on cost and try to become the cost leader, they can have very

dis-tinctive products and strong brands and compete by being different (a strategy of

differentiation), or they can compete through innovation Competing on cost means

Fig 1.5 Porter’s five forces model

using IS to reduce production, administration and marketing costs Increasingly

it means working with suppliers and customers as well by making linkagesbetween the systems in the different organisations The closer customers andsuppliers become tied together – through just-in-time supply systems – then themore costly it becomes for customers to switch suppliers Competing by differ-entiation means creating a distinction between your products and services andthose of your competitors Information systems are used to offer superior servicethrough order entry and order enquiry systems and total customer care systems.The introduction of customer affinity cards and the associated systems by super-markets is an attempt to differentiate themselves from their competitors and tomake it possible to market special offers to specific groups of customers Compet-ing through innovation is about finding new ways to deliver customer satis-faction It could be through the use of IS in research and development or inbringing products and services to the market quickly This is very much the casenow in the financial services industry with the constant introduction of new sav-ings and pension products

In this section then we’ve looked at data and information and seen how dataneeds to be processed or transformed to become valuable information The char-acteristics of useful information have been reviewed, and we’ve seen in outlinehow information and the use of information systems support the strategy of anorganisation But we need to end with some cautionary notes Research pub-lished in 1998 by the Oxford Institute of Information Management at TempletonCollege questions some of the generally accepted ideas about IT and competitiveadvantage According to this research several questions are being asked, and atleast two of them are relevant to us at this point:

• Are claims about IT and competitive advantage overdone?

• Is it beginning to run out of corporate support?

There is general acceptance that IT is being used as a competitive weapon, butnow there is doubt about whether marked competitive advantage is beingachieved Arguments are that ‘We are different from the well known exemplarcases’ and that in any case competitive advantage is short lived: ‘We introduced

a new product and within weeks someone copied it’ The key issues here are thatcompetitive advantage comes from a detailed analysis of the special forces faced

by an organisation and an integrated response over several years The mentation of IS to support strategic change doesn’t deliver immediate knockoutblows Short-lived examples of competitive advantage are well known: super-market affinity cards are an example The actions that turn them into sustainablecompetitive advantage include restricting competitor actions and giving greateraccess to customers In the case of supermarket cards these appear to have beenabsent

imple-In conclusion, then, information management and the use of IT/IS for petitive is a lively topic The fact that it now appears in business and computingdegree courses as well as in business and computing journals is a good example

com-of its growing importance We return to it again in Chapter 12

1.5 Successful Systems

When the new system you’ve worked on is implemented and running regularly,and you’re assigned to another project, how will you know if you did a good job?How will you know if you’ve helped to produce a successful system? A typicalquestion often used in analysis and design examinations asks:

You’re called in to evaluate the effectiveness of a recently implementedsystem What criteria would you use?

Leaving aside project management considerations such as implementation totime, cost and quality you could ask the following questions:

• Does the system achieve the goals set for it? Some of these will be operationalrunning goals concerned with performance, some will be system goalsconcerned with the production of outputs, and some will be business goalsaddressing the purpose of the system development

• How well does the system fit the structure of the business for which it wasdeveloped? The new system will no doubt have been developed based on

an understanding of the then present structure of the organisation and someappreciation of how it might change in the future However, it must not be

an ‘albatross system’ that hangs around the organisation’s neck limiting itsmovement and freedom to reorganise Systems should be designed in aflexible way so that they can be changed to meet changing businessconditions

• Is the new system accurate, secure and reliable? There will be basic

requirements for financial control and auditing, but the system should also

be robust so as to continue in operation with degraded performance duringpartial failure Security from unauthorised access has also now becomeincreasingly important with the growth in the development of tactical andstrategic information systems

• Is the system well documented and easy to understand? Increasingly largeproportions of the budget of system development departments are being used

in the maintenance and updating of existing systems The biggest single way

of limiting these expenses in the future is to take account of it when we designtoday the systems of tomorrow

This ‘single-system’ view may help to identify the characteristics of successfulsystems, but it doesn’t give a sufficiently wide framework for our analysis Weneed to begin with an overview of the organisation as a whole We can, for exam-ple, see the organisation in systems terms, operating within its environment andmade up of a series of subsystems A representation of this that has been widelyused is shown in Figure 1.6

It shows an industrial organisation with subsystems for:

• marketing and purchasing: these are the main links with the environment asrepresented by customers and suppliers It’s important to recognise, however,

Successful Systems 15

Fig 1.6 The organisation as a system

that the environment also interacts with the organisation through legislation,social pressures, competitive forces, the education system and political

• support systems: these are shown as the accounting, personnel and

management control subsystems

For this organisation to work effectively it has to make good use of information,

so the need arises for information systems that collect, store, transform and play information about the business We can represent this information systems

dis-Successful Systems 17

Fig 1.7 An hierarchical view of systems

structure in two ways: either in a non-hierarchical way showing each subsystem

on the same level, or in an hierarchical way where some systems sit on top ofothers This multilevel view is often more helpful as it shows the different levels

of control, the different data requirements, and a different view of the isation of each system A typical way of representing this structure is as if thesystems are arranged in a pyramid as shown in Figure 1.7

organ-At the top level are strategic systems and decision support systems that

informate the organisation Informate is a term that, in brief, means what happens

when automated processes yield information that enables new competitiveadvantage to be gained At this level systems contribute towards the formulation

of corporate policy and direction These are concerned almost entirely withfinancial information and other data that show the health of the organisa-tion Strategic systems use information from lower-level internal systems andexternally obtained information about markets, social trends and competitorbehaviour

Underneath strategic systems lie managerial or tactical systems that are cerned with the monitoring and control of business functions There is therefore

con-a regulcon-ar supply of dcon-atcon-a from the dcon-ay-to-dcon-ay opercon-ationcon-al systems, which con-aremanipulated to provide the management information that these systems typic-ally produce Systems requirements here are for timely, useful and effectivelypresented reports that enable middle managers to run their departments anddivisions effectively Here also we see the need for systems that can respond

to a changing array of ad hoc queries about why the results look the way they do.The operational systems level is concerned with the routine processing oftransactions such as orders, invoices, schedules and statements They help theorganisation to ‘do what it does’ – make parts, distribute products, manageproperty They are not concerned with changing the way the organisation works

We can now see that the organisation can be viewed as a system, and that theinformation systems that support it can be strategic, tactical or operational This

is summarised in Box 1.1

There is one more view that we must see before we can leave this topic It isconcerned with the evolution of information systems A useful model here andone widely used is the Gibson–Nolan four-stage model of:

• initiation;

• expansion;

• formalisation;

• maturity

Let’s consider an organisation moving through this model for the first time

Dur-ing the initiation phase the first computer-based systems to be developed are

those best suited to the new technology These projects are almost always at thebottom of the systems pyramid and involve the repetitive processing of largevolumes of transactions They often begin with accounting systems Following

experience here the organisation enters the expansion stage and seeks to apply the

new technology to as many applications as possible This is the honeymoonperiod for the IS department until one day a halt is called to the ever-growing ISbudget and the introduction of development planning and controls signals the

Box 1.1 The three-level systems summary

Strategic Provide information to managers to enable them to make

systems better-informed decisions

Support decision-makers in situations that are poorlystructured

Used to establish plans for the introduction of new businesslines or their closure

Need greater flexibility to be able to respond to constantlychanging requirements

Tactical Use stored data from operational systems

systems System outputs are well defined as managers can generally

identify the factors influencing decisions that they will have

to make

Usually concerned with the management and control ofdepartments or functions

Operational Transaction based

systems Handle the routine business activities of organisations

Often the first systems to be automated as they provide theraw data for higher-level systems

Successful Systems 19

start of the formalisation stage It is here that the need for information surpasses

the need for data and where the organisation begins to plan its way from amixture of separate data processing systems towards a more coordinated andintegrated approach Corporate recognition of the need for integrated systems

is the characteristic of the maturity stage Here we see the use of open system

architectures, database environments and comprehensive systems planning.There is a final complication Organisations don’t go through this model once.There is no final nirvana of maturity, of fully integrated systems all talking toeach other and producing exactly the strategic information needed by top man-agement The organisation or its component parts can be at different stages in the

model at the same time Having arrived at the formalisation stage with large mainframe computers, many organisations were plunged back into initiation and expansion with the arrival of personal computers New technology, the develop-

ment of new applications software packages or significant price reductions ineither hardware or software all throw organisations back into the initiation stage

At the beginning of this section we asked a question about evaluating theeffectiveness of a recently implemented system We conclude with some gen-eral ideas on wider aspects of this topic Surveys in the late 1990s have shownthat the issue of measuring the benefits of IT investment has been one of the mainconcerns of senior IT/IS management It also appears that, although attentionwas given to cost–benefit analysis at the beginning of the project, much lessattention was given to post-implementation review, and there was consequentloss of opportunity to learn from past project success or failure Consequentlythere has been some new work done on sample projects to evaluate the out-comes throughout the lifecycle to give a better understanding of how best tomanage projects to achieve the desired results

Taking this together with some of the earlier comments about the use of IS/IT

to deliver and sustain competitive advantage, perhaps the best that can be said atthe moment is that the jury is still out

1.6 The Role of the Analyst and Designer

Earlier in this chapter we said that systems analysts and designers change theworld In discussing the role of the analyst and designer we should thereforebegin with a consideration of this change-making initiative We have a problem,however, in that analysts and designers are not always the same person This isillustrated by a diagram produced by G.M Weinburg (Figure 1.8)

It shows very clearly down the centre what he regarded as the analyst’s role.The designers and implementers do quite different things, which are not speci-fied, and the analyst is the only link with the users We also see the difference inrecruitment specifications, where those for analysts ask for applications and busi-ness experience and those for designers specify software environment and hard-ware architectures Life is complicated further by the many sorts of analyst anddesigner titles, such as business analyst, applications analyst, database designer,network designer, database administrator, infrastructure manager and systemsmanager They all draw on the skills that were once the exclusive territory of the

Fig 1.8 The role of the analyst (Weinburg)

systems analyst Let us therefore take a step up from this detailed picture thatprevents us from seeing ‘the wood for the trees’ We can recognise a set ofattributes at a general level that all analysts or designers should possess, what-ever their job title As a minimum we should expect our analysts or designers:

• to uncover the fundamental issues of a problem These might be the

bottlenecks in a business system or the logic of a file-processing module;

• to be able to prepare sound plans and appreciate the effect that new datawill have on them, and replan appropriately;

• to be perceptive but not jump to conclusions, to be persistent to overcomedifficulties and obstacles and maintain a planned course of action to achieveresults;

• to exhibit stamina, strength of character and a sense of purpose essential in aprofessional specialist;

• to have a broad flexible outlook, an orderly mind and a disciplined approach,

as the job will frequently require working without direct supervision;

• to possess higher-than-average social skills so as to work well with others, andthe ability to express thoughts, ideas, suggestions and proposals clearly, bothorally and in writing

Many designers with very special technical skills and knowledge have in the pastsought refuge in their specialisms from the need to exhibit many of the skills andattributes in this list, but in our view the complete list has validity for all analystsand designers whatever their job title

Finally, in this section are two more light-hearted views about the role of the

systems analyst The first is an American view based on Modell’s A Professional’s Guide to Systems Analysis, in which he lists the many roles of the analyst, some of

which are:

• Detective A detective, whether private or official, is one whose primary task is

to uncover the facts of an event and to determine responsibility for the event

• Puzzle solver The puzzle solver is one who either puts things together from

component pieces or determines solutions from clues and hints

• Indian scout An Indian scout is one who is usually the first on the scene and

who looks for hidden dangers or for the correct path through the wilderness(of the corporate environment) The Indian scout may also be the first one tofind hidden dangers, and may draw the first fire

Other roles he proposes include artist, sculptor, diagnostician and reporter.Perhaps he was intending to entertain as well as inform his students when heprepared his list Someone who clearly believes in doing both is Roy Tallis,

a longtime friend and fellow trainer, who used to offer his students someAlexander Pope (1688–1744) with their systems analysis when he describedthe analyst as ‘Correct with spirit, eloquent with ease, intent to reason, or polite

to please’

1.7 Ethical Considerations

Mostly in this book we are concerned with getting the job done, with the calities of analysis and design In this section we’ll consider some of the ethicalconsiderations of working as a systems analyst We suggest that you need to con-sider three issues:

practi-• morality;

• ethics;

• professionalism

Morality is not about hacking or computer fraud or the misuse of computers.

There is legislation to cover these criminal acts They may or not be immoral, but

Ethical Considerations 21