Journal of Information Technology (1992) 7,189-201

Information technology as a service operation

TONY RANDS
Templeton College, Kennington, Oxford OX] SNY, UK

Research into service management has revealed concepts and ideas which may be usefully applied in other
management settings. In order to evaluate their potential relevance to managing information technology (IT), several
aspects of service management are explored, including the nature of service products, customer-supplier contact,
the design of service delivery systems, and measuring and achieving service quality.
Someideas have already been applied to IT, but others could also be relevant. These include examining contacts
between users and IT specialists, using process flow charts to design IT services, and employing service quality
measures in IT management. A general model is proposed for managing IT service quality; the application of this
model at the strategic level of firms is examined.

Introduction (Johnson and Morris, 1985;Bruggeman etal., 1988).Now

it appears these works have revealed new perspectives on
Applying service concepts to IT
management in other settings; recent papers by Kyu Kim
With growth in the service sectors of western economies, (1990) and Galloway and White (1989) suggest the
an awareness has developed that providing a service is, in relevance of service concepts to managing IT should be
many respects, similar to supplying goods. Levitt (1972) taken seriously.
argued that techniques developed in manufacturing were One reason insights from services may be relevant to
relevant to services, while Sasser et al. (1978) applied management in other settings could be a growing
manufacturing thinking to designing service delivery acceptance that few products consist entirely of goods or
systems, e.g. planning capacity levels. Possibly nowhere services (Sasser et al., 1978; Shostack, 1982; Lewis, 1988;
have techniques developed for manufacturing goods been Hill, 1991). Johnson and Morris (1985) see a firm's
applied more rigourously than in fast-food franchising, products lying on a continuum of physical goods (what
where McDonald's has become synonomous with a you buy) and processes (how it is provided). They position
service based around production principles (Sasser et al. car manufacturers' at one extreme, orientated towards
1978). physical goods, with undertakers at the opposite, being
While similarities were recognized between supplying service or process orientated. On this continuum, it may
goods and providing services, several writers, including be argued IT provides mainly a service, rather than a
Millsand Moberg (1982), cautioned against the wholesale physical, product for firms.
use of manufacturing techniques in service industries, This suggests managing IT may be similar to managing
sincemajor differences existed. For instance, while many other services which utilise equipment to provide at least
goodsmay be stored, services, such as airline seats, often part of the service product. However, this observation is
have no shelf-life, leading to simultaneous service not new, as, for a number of years, the service aspect ofIT
production and consumption (Sasser et al. 1978;Mills and has been recognised. Rockart (1982) identified "Service
Moberg, 1982; Lewis, 1988, 1990). Also, whereas and the perception of that service by management" as a
customers are usually exterior to goods manufacturing critical success factor in managing information systems.
processes, they are normally integral to service processes Satisfaction, a measure used in service settings, is also
and contribute to service production (Gronroos, 1984; employed to assess IT applications, while, in a study by
Johnson and Morris, 1985; Haywood-Farmer, 1988; Martin (1990), several firms said they would like to
Larsson and Bowen, 1989; Wild, 1989). introduce Service Level Agreements. As mentioned
These differences have led researchers to develop earlier, what is new are recent papers which have
methods of analysis which address service environments. approached IT management using concepts developed
The spotlight has fallen on several aspects of services, principally in service settings.
including quality (Gronroos, 1984; Zeithaml et al., 1988; Research at Templeton Collegesuggests the application
Brogowicz et al., 1990; Lewis, 1990), technology (Mills of service concepts to IT marks an important departure
and Moberg, 1982), delivery system design (Shostack, from previously accepted methods of analysis, which will
1982; Sasser et al., 1978), and management control yield new insights and methods for managing IT. In order
190 Rands

to illustrate the nature of the departure and its potential (Parasuraman et al., 1988) reduced to five. These
consequences, this paper firstly summarises concepts of dimensions, as well as derivatives of the original ten, have
service management, and then examines their possible since been widely accepted as providing a basis for
application to IT, concluding with suggestions of research defining service products. The five are:
directions to investigate applications areas further.
(1) Tangibles - physical evidence of service
Aspects covered include defining service products,
(2) Reliability - ability to perform the promised service
customer contact, designing service delivery systems and
dependably and accurately
service quality.
(3) Responsiveness - promptness and willingness
(4) Assurance - knowledge about client and ability to
inspire trust and confidence
The service product
(5) Empathy - caring, individualised attention
Mismatches between customers' and producers' provided to customers
perceptions
While these dimensions relate service products to
There are other distinctions between service and physical customers' needs, reflecting both physical and functional
products, besides possessing no shelf-life and involving components, Wild (1980) sought to link services to the
the customer in production. Services are usually design of their delivery systems; for this he identified
intangible, being the perceptions of customers about the three broad service categories of his own. These were the
processes of a service provider (Sasser et al., 1978; Specification of the service, its Cost and Timing. Within
Johnson and Morris, 1985; Haywood-Farmer, 1988; each category he defined narrower parameters. As, other
Lewis, 1988). Also services are consumed, rather than than Cost, parallels may be drawn between Wild's
possessed, by customers (Lewis, 1988), consumption on narrower parameters and Parasuraman and colleagues'
a package holiday being the use of facilities and goods, e.g. proposals, it would appear their five dimensions could
the airline and hotel food (Johnson and Morris, 1985). provide the basis both for defining service products and
The intangibility of services means that, in contrast to designing delivery systems.
manufacturing goods where the product is usually well
defined, differences often occur between what a firm
believes it supplies and what customers feel they buy Customer contact
(Johnson and Morris, 1985). This means mismatches may
Classifying services organisations using contact
occur between the service products perceived by
producers and consumers (Zeithaml et al., 1988). As, in The recognition that customers were integral to service
many circumstances, 'the service product is the service production, while service products were customers'
process itself (Johnson and Morris, 1985), designing perceptions about service processes, has led to several
service products consists of developing their delivery analyses of contact between service firms and customers.
systems (Sasser et al., 1978). To avoid mismatches Larsson and Bowen (1989) have argued the nature of
between customers and service producers, therefore, service products depends on how contact is managed.
customers' perceptions of service products should be Contact often occurs through employees (Lewis, 1988),
understood when service delivery systems are designed although, in many services, this has been reduced by
(Sasser et al., 1978). substituting technology, for instance vending machines
(Sasser et al., 1978). The aim has usually been to achieve
consistency and reduce costs, but increased availability
The dimensions of service products
has also been a reason (Sasser et al., 1978).
As services products are generally based on consumers' Although technology might be employed in services,
perceptions, attempts to define them have been based Sasser et al. (1978) contend interactions between
around customer needs. Sasser et al. (1978) saw customers customers and employees may be key to those services, as
as having a Substantive Need for the essential parts of a some human attributes cannot be substituted. Also
particular service, plus other needs which include Zeithaml et al. (1988) argue employee-customer contacts
Control, Trust, Consequence, and Self-fulfillment or create quality. One reason contact with employees could
Status. In order to satisfy these needs, services are be important is customers' contributions to producing
generally composed of physical (what is provided) and services are often crucial to those services (Larsson and
functional (how it is provided) components (Sasser et al., Bowen, 1989), and depend on contact with employees.
1978; Shostack, 1982; Lewis, 1988; Hill, 1991). Larsson and Bowen have called this 'co-production'.
Surveys by Parasuraman et al. (1985) revealed that, in Although customers' input to service production may be
evaluating services, consumers used criteria which fitted important, Larsson and Bowen (1989) argue there has
ten potentially overlapping dimensions, which they later been a tendency in many services to seek to buffer
IT as a service operation 191

customers from the core service technology. By doing High

this, customers are treated "only as consumers, not as Quadrant III Quadrant II
producers" (Larsson and Bowen, 1989). Sequential Reciprocal Service
Due to its importance in shaping service products, Customised Service
Design
Design

customer contact has now been widely accepted as a basis ego car repair,
g. medical care, legal advic e
higher education
for classifying service organisations (Chase, 1981). For Diversity cleaning, gardening

instance, Haywood-Farmer (1988) and Schmenner (1986) of ~-_.

Demand
use the extent of customer-firm contact and human Quadrant IV Quadrant I
involvement to distinguish between airlines, hospital, Pooled Service Sequential
Design Standard Service
schools, doctors, etc. Design
ego banks, airlines
ego laundromats,
car rental
Task interdependencies between consumers and Low
employees.
Low High
Instead of using the extent of contact to classify service Customer Disposition to Participate
organisations, Larsson and Bowen (1989) adopted the
'nature of the task dependency' between customers and Figure 1 Classification of services
employees required to produce the service. For this
they applied a model of Thomson (1%7) for analysing
internal organisational relationships, which associates followed by customers (Wild, 1989), as well as resources.
dependencies of parties to links between the tasks each The latter include the movement and preparation of
undertakes. materials, the deployment and maintenance of equip-
Thomson defines three task dependencies, Reciprocal ment, etc.
where the output of each has to be the input for the other, The inclusion of customers in process flow charts
Sequential where the output of one party alone should be reflects their contribution to service production, and
the input for the other, and Pooled where no direct allows the service received by customers to be identified
relationship exists other than common goals. by the service provider. For instance, charts may show
Larsson and Bowen (1989) propose a grid with axes, where customers have to queue, often for resources to
Diversity of Demand and Customer Disposition to become available, where bottlenecks exist and where
Participate, to categorise firms into four quadrants. A customers contact the service provider. Flow charts may
derivative is shown in Figure 1. In Quadrants II and III, also incorporate service quality checks (Sasser et al.,
where demand diversity is high, services would generally 1978). One limitation is a tendency to focus on physical
need to be customised, whereas in Quadrants I and IV, rather than functional aspects of services; in an attempt to
with low demand diversity, standard services may be overcome this, service elements (Shostack, 1982)and 'fail
offered. In each quadrant, Larsson and Bowen included points' may be included within flowcharts.
the customer-firm interdependency generally needed A simple flow chart for a hypothetical car hire firm is
to produce the service product (this is also shown in shown in Fig. 2. It includes processes undertaken by the
Figure 1). customer, on the car and a number of clerical operations.
The service product perceived by customers is likely to be
processes followed by them, some in contact with the
Designing service delivery systems firm, and is represented by the dotted box in Fig. 2.
Changing these processes, therefore, as well as processes
Process flow charts
to the car, will modify the service product. Technical parts
As in most services, the service process is inseparable from of the service include the car's physical attributes, e.g. its
the service product (Johnson and Morris, 1985), it was size and economy, whilst functional parts include its
mentioned earlier that designing a service consists of cleanliness and availability on collection, attitude of staff,
designing its delivery system (Sasser et al., 1978). Wild etc. By driving the car, the customer participates in
(1980) examines the relationship between specification, producing the service, leading the firm to fall into
cost and timing of a service, and the productivity of Quadrant I in Figure 1.
machine, labour and material resources in its delivery
system.
Front and back office activities
One method of designing a delivery system is to use
process flow charts (Sasser et al., 1978; Shostack, 1982; Although customer contact is important in shaping
Wild, 1989), which represent sequences of operations in services, other activities, which may be undertaken
the system. Flow charts usually incorporate processes remote from customers, are also necessary. In Fig. 2,
192 Rands

therefore, of linking customers with front offices, and

front to back office; Larsson and Bowen (1989) refer to
these as coordinating mechanisms. In Figure 2, Pay Bill and
Assign Car are coordinating mechanisms linking
customers with the front office and front to back offices
respectively.
Larsson and Bowen (1989) have argued that 'the
coordinating mechanisms required in a delivery system
are dictated by the task dependency between customers
and the firm necessary to produce the service'. This means
mechanisms for linking consumers to front offices and
front to back offices should be determined in the light of
the requiredservice and the dependency between clients
and employees. For example, with firms in Quadrant II in
Process Flow
Figure 2, coordinating mechanisms linking customers to
~~--~----=>

Data FloW8
front offices should reflect a Reciprocal task dependency,
Customer Clerical Car
and ensure the output of each becomes the input for the
Figure 2 Car hire process flow chart other. Meanwhile, for firms in Quadrant I, mechanisms
should ensure outputs from customers become inputs to
several operations to the car fall into this category. To the front office.
distinguish customer contact from other operations, they
are usually known as front and back office activities
respectively (Chase, 1981; Larsson and Bowen, 1989; Managing service quality
Hill, 1991). One result of substituting technology for
Service quality measures
employee contact has been to increase the importance of
back office activities (Sasser et al., 1978), for example Service quality is linked to customers being satisfied with
servicing teller machines in the case of banks. the services they receive (Lewis, 1988), and has been
Deciding whether activities should be front or back defined as the degree of fit between customers'
office is normally integral to designing service delivery expectations and perceptions (Gronroos, 1984; Lewis,
systems. Larsson and Bowen (1989) have suggested the 1988; Zeithaml et al., 1988; Silvestro et al., 1990).
division between front and back office work should be However, Gronroos (1984) distinguishes between
determined by 'the nature of the task dependency between technical and functional quality, reflecting the physical
consumers and employees'. In some situations, it may be (what is provided) and process (how it is provided) aspects
unclear whether an activity ought to be front or back of services (Sasser et al., 1978; Shostack, 1982; Lewis,
office. For instance, in Figure 2, Assign Car may be 1988; Hill, 1991).
regarded as either front or back office, depending on the Gronroos (1984) argues technical quality is necessary,
required customer involvement. but not sufficient to achieve customer satisfaction, and
Deciding whether an activity should be front or back suggests overall quality also involves attaining functional
office often influences other factors, such as employee quality. Brogowicz et al. (1990) believe the two often
skills requirements. Each usually demands different cannot be separated, as high technical expectations may
attributes (Lewis, 1988), front offices requiring lead to high functional expectations. They argue that
interpersonal skills, while back offices require technical functional quality is easier for consumers to evaluate than
skills (Zeithaml et al., 1988; Larsson and Bowen, 1989). If technical quality, but may be more difficult for the
Assign Car in Figure 2 were front office, it would demand supplier to measure.
different personal skills than if it were back office. In order to measure service quality, Parasuraman et al.
(1988) developed a tool, SERVQUAL, which recorded
consumers' views about twenty two aspects of services,
Coordinating mechanisms
each related to one of the five service dimensions
Several procedures in Figure 2 may be performed described earlier. Consumers were asked for their
exclusively by the customer or within the back office; 'expectation before and perception after' they received the
hence Drive Car is undertaken by the customer, while service, but not their ranking of the importance of each
Initial Check and Inspection are purely back office aspect. Carman (1990) concluded SERVQUAL was
activities. However other procedures cross boundaries, reliable, but recommended consumers also be questioned
either between customers and the firm, or between the about the importance of each aspect, consistent with
front and back office. These procedures have the effect, services being composed of central and peripheral
IT as a service operation 193

elements (Sasser et al., 1978; Brogowicz et al., 1990). service is unlikely to satisfy customers' needs, causing low
Carman also questioned whether the five dimensions were quality.
generic, or could be amended for services other than
retailing, where SERVQUAL was originally developed.
Achieving quality in customised professional or
Several studies have been made using derivatives of
specialist services
SERVQUAL. Lewis (1990) assessed customers' views of
some major UK banks, building societies and retailers. In Figure 1, Quadrant II contains firms which experience
Meanwhile, Kierl and Mitchell (1990) adapted high demand diversity from customers with a strong
SERVQUAL to measure the service quality of six firms disposition to participate in producing the service.
supplying basic industrial products, perceived by 85 Larsson and Bowen (1989) argue customers of these firms
senior purchasing managers. Their questionnaire often have unique and complex problems requiring
included the importance of each aspect, and the results specialist expertise from employees to customise
showed the gap profile across the service dimensions for individual solutions. To produce this service, Fig. 1 shows
the six firms. Kierl and Mitchell concluded with a map of task dependencies between customers and employees
how each firm fitted consumers' needs. generally to be Reciprocal (Larsson and Bowen, 1989).
This means contacts between employee and customers
should be managed to ensure the outputs of each are the
inputs for the other, reflecting the 'need for the customer
Designing quality into delivery systems
to provide information for adequate problem solving
The simultaneous production and consumption of throughout service production'. This close linking should
services creates obstacles to inspecting their quality lead to 'interactive service production' (Larsson and
(Sasser et al., 1978). This means service quality may best Bowen, 1989). Many professional and specialist services,
be achieved by designing it into the delivery system for instance legal advice and medical care, fall within
(Johnson and Morris, 1985; Sasser et al., 1978). Sasser et Quadrant II (Larsson and Bowen, 1989), because
al. describe a hotel, where complaints about oversights in consumers require expertise, and Reciprocal dependency
preparing rooms, e.g. burned out bulbs, were not solved with employees is necessary to produce the customised
by regular inspections. They were eventually reduced by service.
standardising procedures for preparing rooms. However, in professional services, power distribution
In designing quality into delivery systems, a between clients, professionals and service managers
relationship between service level and service cost to influences the way operating decisions are taken (Harvey,
customers should also be recognised. Sasser et al. (1978) 1990), which in turn impact the quality and quantity of the
identified service level-price combinations, acceptable to service provided. A number of factors influencing power
both customers and producers, and others where the distribution are external. The professional's power,
service level was either too high or low for the price paid represented by autonomy, namely the "absence of ...
by customers. This made the services unprofitable for pressure from clients, from the organisation or from any
producers, or unacceptable to customers. other person" (Harvey, 1990), is partly determined by the
Previous discussions suggest a number of factors are prestige of, and supply and demand for, the profession.
important in achieving service quality. Among these are: Clients' power is partly shaped by competition for their
business and their knowledge of the profession (Harvey,
(1) Front-back office distinctions: a failure to make the
1990). Meanwhile internal factors influencing power
correct front-back office distinction for an activity,
distribution include procedures which allow professionals
e.g. Assign Car in Figure 2, may lead to poor service
to act as gatekeepers to the service.
quality. Contact personnel may lack sufficient
For clients requiring a customised specialist service,
interpersonal skills (Zeithaml et al., 1988) to
therefore, a potential dichotomy exists between the
interact with customers.
Reciprocal dependency with professionals needed to
(2) Coordinating mechanisms should reflect customer-
produce the service, and power distribution between
employee dependency: coordinating mechanisms
clients, professionals and service managers. This may
within a service delivery system should be
allow professionals sufficient autonomy so client contacts
compatible with the dependency required to
do not reflect the Reciprocal task dependency demanded
produce the required service product (Larsson
by the service. As a result, professionals may interpret
and Bowen, 1989). Mechanisms should ensure
clients' requirements differently from clients themselves,
consumers contribute to producing services.
creating service mismatches.
The latter suggests that, when the processes for linking This suggests that, to achieve quality, internal
clients, front and back offices are inappropriate to the task procedures of organisations providing customised
dependency required to produce the service, the resulting specialist services should manage the power distribution
194 Rands

in a manner which permits the Reciprocal task between earlier methods of measuring User Satisfaction
dependency, necessary for the service, to function. Front with information systems, and proposed that service
office activities should be the main focus for coordinating concepts provided a conceptual base. Using these, Kyu
mechanisms (Larsson and Bowen, 1989), with links Kim defined User Satisfaction as a function of the
between front and back offices supporting these. Front difference between users' expectations and perceptions of
office procedures should ensure a continuous an IT application, and identified three gaps within the
transmission of information between clients and Software Development Cycle. These were gaps in
employees. determining information needs, designing and installing
the information system, and delivering the information
service. Kyu Kim's gaps are similar to Zeithaml et al.
Models for managing service quality
(1988), explained earlier, but do not correspond exactly.
For firms to manage service quality and minimise gaps By applying ideas derived from service environments, I
between consumer expectations and perceptions, believe both works have added to our understanding ofIT
Zeithaml et al. (1988) identified four sources for these management over more traditional approaches. Despite
gaps. They attributed these sources to mis-perceptions this, previous discussions suggest other service concepts
which occur during the specification, design and might also be relevant, and provide additional insights
installation of service delivery systems. into defining IT service products, managing contacts
The sources identified by Zeithaml et al. were: between users and IT specialists, and achieving quality in
IT services. These are explored in the remainder of the
Gap 1 - difference between consumer expectations and
paper.
management perceptions of consumer expectations.
Gap 2 - difference between management perceptions of
consumer expectations and the service specification.
Gap 3 - difference between the service specification and The IT service product
the service actually delivered. It was argued in the Introduction that, for firms, the
Gap 4 - difference between the service actually consequence of deploying IT must be the receipt of
delivered and what is communicated to consumers services, as IT installations and personnel in themselves
about the service specification. do not constitute the IT product. This may mean that
Zeithaml et al. (1988) proposed a model for achieving determining the nature of the service product provided by
service quality, which sought to manage these four gaps a firm's IT resources should be linked to benefits it is able
during the design and implementation of service delivery to gather through their use. However, benefits may be
systems. The model, therefore, provides a framework for perceived at a number oflevels; for instance, at the lowest
managing service quality. Brogowicz et al. (1990) later are those derived from individual IT applications, whilst,
modified Zeithaml and colleagues' model to allow for, at the level of the firm, lie overall benefits obtained from
amongst other things, technical and functional aspects of its complete range ofIT resources, which includes people,
service (Gronroos, 1984). The proposed framework for software, equipment, etc.
managing quality in IT services is a derivation of these
models.
Components of the IT service product

Applying service concepts to IT management This discussion suggests the overall service product
provided to a firm by its IT resources could be perceived
Earlier papers using service concepts as the combined outcome from a portfolio of separate IT
It was argued earlier that adopting service concepts could service components, possibly occurring at different levels,
be an important departure from accepted methods for from specific areas to the level of the firm. It is proposed
analysing IT management, which may herald fresh here that individual IT service components in the port-
insights and new approaches. Galloway and White (1989) folio will tend to fall into one of the four quadrants
produced one of the first papers using service concepts, suggested by Larsson and Bowen (1989), with the total IT
which considered IT development projects in terms of service product the combined outcome of the service
front-back office activities. They argued that, whereas component mix. A suggested portfolio of service compo-
Design, Programming and Testing are often treated as nents comprising the overall service product provided by a
back office, they should also be front office. Skills firm's IT resources is shown in Figure 3. Should the mix
required for contact with users are often not possessed by of service components within the portfolio of Figure 3
programmers and analysts in typical DP departments. change, so the total IT service product provided to the
Meanwhile Kyu Kim (1990) examined inconsistencies firm would be modified. Components comprising the
IT as a service operation 195

portfolio which forms the overall IT service product High Quadrant III Quadrant II
include IT strategy formulation, equipment maintenance, Sequential Reciprocal Service
and help desks. Customised Service Design
Design
In Figure 3, applications development has been IT Strategy Formulatio
included as a service component. This is because the Data Base Maintenance Application Developmer t
product of the applications development process within Extent Help Desk Package Selection
Applications Updating
the IT function should not be conceived as a delivered of
Customising Quadrant IV Quadrant I
system, for users may feel this to be only part. They might
also believe the product includes the way the system is Pooled Service Sequential
Design Standard Service
delivered, e.g. its timeliness, difficulties in installation, Design
etc., and other factors, such as training, helpfulness of IT Equipment Maintenance End-User Development
Network Operation
staff, etc. These items are not physical and have parallels Low
Support

with the functional (the manner of delivery), rather than

the physical (what is delivered), aspects of services Low High
(Gronroos, 1984). Thus the product of the IT Function's User Disposition to Participate
development processes may also be perceived as a service, Figure 3 Components within an IT Service Product
similar to helps desks, composed of physical elements, i.e,
delivered systems, manuals, etc., and functional parts,
such as those just discussed. Although this paper has claimed that service concepts
In Figure 3, each service component has been assigned provide additional insights into managing IT to traditional
to the quadrant felt most appropriate. For instance, the methods, it is recognized here that using service concepts
application development service is placed in Quadrant II, is unlikely to remove this basic impediment to assessing
as customizing is usually high and users often have a IT effectiveness through direct measurement, caused by
strong disposition to participate, while the help desk the influence of uncontrollable and unmeasureable factors
servicehas been placed in Quadrant III, as users may have on firms' performance. Thus it would appear that con-
a wide range of queries and not be disposed to participate sidering IT from a service perspective provides alternative
in creating the service. On the other hand, if the end-user surrogate measures of IT effectiveness to those mentioned
development support service provided were standard, e.g. earlier. In this case, the surrogate measure would be the
confmed to assistance with creating Lotus models, then quality provided by individual service components within
this would fall in Quadrant I, as user disposition would be the IT service product, typified by the portfolio shown in
high. Finally, the network operation service has been put Figure 3. The implicit assumption would be that, if the
in Quadrant IV, as this is likely to be standard, with user mix of service components within the portfolio were con-
disposition low. sistent with the required IT service product, and the
quality of each component were sufficient, then this
would lead to overall IT effectiveness within the firm.
Service quality as a surrogate measure for
IT effectiveness
Dimensions of IT service components
It was suggested earlier that defining IT service products
may need to be linked to benefits a firm is able to gather Each service component within a firm's overall IT service
through using its IT resources, comprising equipment, product is created by users' interaction with a range of
people, software, etc. However, many uncontrollable and different IT resources, e.g. equipment, software, staff,
unmeasurable factors influence firms, often making it etc. which have been assembled to provide the delivery
impossible to attribute performance improvements system for that particular service component. For
directly to the use ofIT (DeLone, 1988). This creates an example, users consuming the help desk service may
impediment to measuring IT's effectiveness and contri- mainly interact with IT staff located at terminals through
bution, across all levels offirms, through benefits directly telephones, whilst those using the network operation
accredited to the use of IT resources. Instead surrogate service may primarily interact with keyboards and
measureshave been developed (Baileyand Pearson, 1983; screens, and only occasionally with operators.
Miller and Doyle, 1988; DeLone, 1988), some of which With a range of different IT service components
relate to users' reactions. For individual IT applications, potentially available to firms, it would appear advan-
these reactions include users' feelings, e.g. user satis- tageous, therefore, if a consistent method of defming each
faction (Bailey and Pearson, 1983; Doll and Torkzadeh, component could be provided, which could be applied
1988;Ives etal. 1983),and the extent applications are used across the entire range and permit comparisons between
(DeLone, 1988). service components. Also it would seem desirable for the
196 Rands

Improved efficiency
Performance Increased effectiveness

Dimension
r- -Tangibles
I - Physical evidence
_ r-- -Reliability

a er Vice
•
Prov/~
I - Accuracy & Dependebility
- - ~esponsiveness
: - Timing
r -Assurance
& Speed of Reepon.e

~Vice
I .- Relevance and Trust

L - -Empathy
- Attention to User

( C;;~onent
livery Sys m
<,

?
Intellectual Products
Technical ~ ./ -, ~ (Concepts, ideas,
Architecture Business Understanding)

Systems Staff
Architecture Service Infrastructure
& Portfolio (eg Staff distribution,
project request mechanisms)

Figure 4 Model of IT Service Components

method to reflect both their technical (what is provided) service may be comprised principally of IT managers and
and functional (how it is provided) aspects (Gronroos, documentation.
1984). For in each service component, functional aspects I believe that viewing IT services as multi-dimensional
may be as important as technical in determining quality. using the five dimensions of Parasuraman et al. is a radical
For example, help desk service quality may depend, not departure from earlier approaches to IT, which may
only on equipment installed, but helpfulness and significantly increase our insight into IT management.
competence of staff. Traditional approaches have tended to treat IT in uni-
This need to reflect the technical and functional aspects dimensional terms, with the possible result that the
of each service component implies that the chosen method impact of physical aspects on IT effectiveness has been
of definition should regard IT service components as over-emphasized. One example of this could be user satis-
multi-dimensional (Parasuraman et al., 1988). With this faction, which has often been measured using a single
in mind, it is proposed that, subject to reservations scale, with the result that several tools developed (e.g.
expressed by Carmen (1990), the service components of Bailey and Pearson, 1983; Montazemi, 1988; Doll and
the IT service product should each be measured using the Torkzadeh, 1988; Ives et al. 1983) possibly have a
five dimensions of Parasuraman et al. (1988). A general technical bias in questions put to users.
model which incorporates these dimensions is given in
Figure 4. Here a single service component within the
portfolio which forms the IT service product is shown. User/IT staff contacts
User performance, stemming from the provision of the Earlier descriptions of the perspective provided by service
service component, is influenced by service quality, and concepts suggest contact between users and IT staff may
provides benefits, small or otherwise, to the firm. The be key in creating the required quality for each service
service component in Figure 4 is provided by its own component of the IT service product shown in Figure 3.
individual component delivery system, a sub-system of It is possible the following aspects of contact may be
the overall IT delivery system, consisting of physical applicable.
resources (e.g. equipment, software, staff), service
infrastructure, intellectual products, etc. In the case of the (I) Task dependencies required to produce IT service
help desk service, the delivery system may comprise components: Figure 3 has demonstrated that
telephones, staff, equipment, documentation, etc. Mean- service components of the IT service product
while the delivery system for IT strategy formulation may be differentiated according to the degree of
IT asa service operation 197

customizing required, and users' disposition to Figure 3, the main locus should be linking between
participate in service production. This suggests front and back office IT staff.
task dependencies between users and IT staff (4) Managing user participation: Bowen et at. (1989)
should match the quadrant to which each service have highlighted the importance of service pro-
component is assigned. For instance, providing viders managing customer participation during
the application development service demands service production. As the required user partici-
reciprocal relationships between users and ITstaff, pation in producing each IT service component
with outputs of each becoming inputs to the other. identified in Figure 3 is likely to be different, this
The IT strategy formulation service would also means managers of the IT function must provide a
appear to require reciprocal relationships. On the range of user participation patterns, each consistent
other hand, providing the help desk service for with its relevant service component.
users with IT knowledge, e.g. end-users (Doll and In the case of the application design service,
Torkzadeh, 1988), may require sequential task Figure 3 has suggested users have a high disposition
dependencies between them and IT staff. to participate, and relationships with IT analysts
Whichever task relationship may be required to need to be reciprocal. This may mean that, in order
create a particular IT service component, Larsson to manage user participation satisfactorily, the
and Bowen (1988) have highlighted the importance software development cycle used by the IT function
of managing consumer contact with staff, so con- should include processes for user role-making, as
sumers become part of service production. This well as procedures to ensure users possess relevant
would suggest that, however slight user/IT staff skills (e.g. IT knowledge) to contribute towards the
contact may need to be to produce a particular production of the required system (Bowen et al.,
service component in Figure 3, e.g. the data base 1989). Without these, reciprocal relationships may
maintenance service, that contact should be not be developed between users and IT staff.
managed so users contribute to service production. (5) Managing user/IT staff power distribution: Power
(2) Front/back office IT activities: Gallowayand White distribution between users and IT staff may be
(1989) have already drawn attention to front and shaped by a number of factors, such as the nature of
back office aspects of managing the IT function IT staff specialisms, IT staff shortages and turn-
during applications development projects, namely over, resource allocation procedures which may
the application development service in Quadrant II allow IT staff to act as gatekeepers, users' reliance
of Figure 3. Their analysis has highlighted the need on the internal IT function to provide services, etc.
to increase the inter-personal skills of analysts and It was shown earlier that the management of this
programmers so contacts with users, namely front distribution could be critical in the case of
office activities, produce a satisfactory service Quadrant II services.
during development projects. However Figure 3 The principal aim of this paper is to introduce
suggests other service components would require a service concepts in the context of IT, and it is
different balance between front and back office outside its scope to pursue specific topics in depth.
activities. For instance, in the case of the network However, a hypothetical example of power distri-
operation service, many activities may be back bution in IT services falling in Quadrant II of
office, with few front office. Figure 3 may illustrate its importance. Here, one
(3) Co-ordinating mechanisms: Larsson and Bowen's of the IT service components is the application
(1989) proposition, that coordinating mechanisms development service. If the power distribution in
linking consumers, front and back offices should operation allowed IT staff providing this service,
reflect task dependencies required to produce a in this case specialist analysts and programmers,
service, demonstrates how users, and front and considerable autonomy, a dichotomy could exist as
back office IT staff should be linked to each other. the required reciprocal task dependency may not
For instance, to provide a satisfactory application function in practice. Instead of procedures co-
development service where reciprocal task ordinating users' interactions with front office IT
dependencies are necessary, the main locus should specialists being the main locus, and designed to
be procedures co-ordinating interactions between ensure 'mutual adjustment' through 'large, loosely
front office IT staff and users to ensure 'mutual specified scripts', those procedures could become
adjustment' through 'large, loosely specified secondary to back office activities such as pro-
scripts' (Larsson and Bowen, 1989); meanwhile gramming. Moreover, the procedures could also be
there should be limited decoupling between front in the form of 'standardized, tightly specified
and back office IT staff. On the other hand, to scripts' (Larsson and Bowen, 1989). Such pro-
provide the help desk service in Quadrant III of cedures, therefore, would be more appropriate to
198 Rands

Quadrant IV services than Quadrant II (Larsson To achieve quality in each IT service component,
and Bowen, 1989), as IT specialists, rather than service concepts suggest:
users, would tend to dictate the specification,
(1) The use of SERVQUAL-type tools to assess IT
timing and other aspects of systems.
service quality: IT versions of SERVQUAL which
It may be argued that applying service concepts to IT measured quality along five dimensions could
can lead to increased conceptual clarity about contacts provide sounder surrogate measures for IT
between users and IT staff compared with earlier effectiveness than single scale measures. Various
approaches. For instance, in the case of user involvement, items would be included for each dimension in
or participation (Barki and Hartwick, 1989), during Figure 4, and users asked for their expectations,
information systems development, various studies have perceptions and, possibly, importance (Doll and
come to mixed conclusions as to whether its impact was Torkzadeh, 1988; Carman, 1990)for each item.
positive or otherwise (Ives and Olson, 1984). However, an In the case of assessing quality of the appli-
examination of the studies reviewed by Ives and Olson in cations development service, SERVQUAL tools
the light of the service concepts, just discussed, reveals could incorporate items already available from
potential difficulties in their approach, which may have existing user satisfaction measures, e.g. Miller and
made mixed conclusions inevitable. For service concepts Doyle (1987). However additional items may also
demonstrate that, on its own, user participation in systems be required, possibly linked to dimensions which
development cannot guarantee improved systems, unless highlight functional aspects of the applications
it occurs in circumstances which allow reciprocal relation- delivery service, such as Empathy.
ships with IT specialists. Should circumstances create (2) Designing quality into IT delivery systems: Service
other relationships, for instance sequential, due, say, to concepts suggest that quality of all IT service
the power distribution in operation or a systems develop- components is best achieved by designing it into the
ment cycle which failed to define users' roles adequately, delivery system of each (Sasser et al., 1978). This
then user participation would not necessarily lead to the has already been recognized in certain areas, for
application development service in Figure 3 producing instance the applications development service,
satisfactory systems. Mixed conclusions across the studies where Rahman (1987) has argued that poor soft-
reviewed by Ives and Olson could have stemmed from a ware quality cannot be avoided by final inspection
failure in those studies to identify whether the user alone; processes which build quality into appli-
participation they recorded took place in circumstances cations must be incorporated into the systems
which permitted reciprocal relationships. development cycle. Rahman is supported by a
report for the DTI in the UK (Price Waterhouse,
1988) which said 'it is more important to build
Measuring and achieving IT service quality quality through the development process, rather
than identifying faults by testing the finished
Users and IT staff are likely to develop different
product'. However, this message is, apparently,
perceptions about each IT service component, possibly
not widely appreciated. A survey of over 400 firms
leading to serious mismatches. It may be easier for users
by Vickers et al. (1987), covering a cross-section
to assess functional rather than technical quality of an IT
of industries in the US, showed less than 32% had
service, but difficult for IT staff to measure functional
even formal quality standards for software develop-
quality (Brogowicz et al., 1990); therefore mismatches
ment.
may concern the perceived relevance of technical and
(3) IT service process flow charts: The use of process
functional elements for each service component in
flow charts to design the delivery systems of
Figure 3. individual IT service components may assist in
The framework of Parasuraman et al. (1988) suggests
identifying the service product in each case, by
quality measures for IT service components should be including processes covered by users, as well as IT
based around the five dimensions proposed earlier, there-
staff. Process flow charts may incorporate quality
by indicating both functional and technical performance. procedures, e.g. as proposed by Price Waterhouse
They should also incorporate differences between users'
(1988) for IT applications, as well as coordinating
expectations and perceptions of each service they receive. mechanisms linking users to front offices, and front
Kyu Kim (1990) has used the latter thought to propose a to back offices.
measure for user satisfaction with information systems
applications, namely the applications development Figure 5 shows the process flow chart for a hypothetical
service, but made no suggestions about service dimen- application development service, based around the
sions. traditional systems development cycle. The service
IT asa service operation 199

------------
model by Zeithaml et al. (1988), referred to earlier, which
was later modified by Brogowicz et al. (1990).
Figure 6 attributes gaps between users' expectations
and perceptions of a particular IT service component to
misperceptions which occur during the specification,
design and installation of its delivery system shown in
Figure 4. Perceptions of users' expectations by IT
management provides the basis for service component
specification, i.e. determining the required features of the
service component delivery system. Following this, the
service component delivery system is installed, and is
composed of various IT resources, as shown in Figure 4.
IT management, explicitly or implicitly, passes communi-
prcceee Flow
cation to users about the service component specification,
Data Flows which helps form the service image for that component.
Users IT Specialists
Meanwhile, users' experiences of the service compo-
Figure 5 Application development service flow chart nent delivery system influences their opinion about the
perceived technical service and perceived functional
service. Perceived service is judged partly from these
components, but also from communication to users.
Expected service is a combination of expected technical
product, as perceived by users, is represented by the service and expected functional service, shaped by needs,
dotted box, and is composed of processes involving them, word of mouth, past experience, service component image
i.e. use system, training and accept system, etc. Processes and the technology environment.
and documentation used by IT specialists during the In Figure 6, service component quality is measured by
software development cycle have also been included. the service gap, i.e. the difference between the expected
There are a number of coordinating mechanisms, for service and perceived service, and is a combination of the
instance accept system. technical gap and the functional gap. Sources for the
Figure 5 illustrates how, in the case of the traditional service gap occur during the specification, design and
systems development cycle, users and IT specialists may implementation of the service component delivery
form different perceptions of the service product pro- system. In this framework, four gaps are identified. These
vided, which might lead to serious mismatches. For are:
instance, the latter may believe the product consists of the
Gap I - difference between expected service and
output of the systems development cycle, i.e. the installed
perceptions of user expectations within the IT function.
system, whereas, as mentioned earlier, users may see the
Gap 2 - difference between perceptions of user
product in terms of how the system was delivered, the
expectations within the IT function and service
effectiveness of training, etc. Figure 5 also shows how the
component specification.
service product and its quality may be changed by modi-
Gap 3 - difference between service component
fying processes involving users and IT specialists. For
specification and the installed service component
instance, the addition of 'develop prototype' between
delivery system.
define and design with a feedback for users to 'check
Gap 4 - difference between installed service component
prototype', to incorporate prototyping, would change the
delivery system and communication to users.
service perceived by users, as well as its quality.
This framework may provide the basis for managing the
quality of different IT service components within the
A proposed framework for managing IT service portfolio of Figure 3, for instance, help desks, end user
quality development support, etc. A version of Figure 6 has
already been applied to the application development
In order to bring together the ideas which have just been service by Kyu Kim (1990), who traced differences
discussed, a general framework is proposed, which may be between users' expectations and perceptions to three gaps,
applied to each IT service component in Figure 3. This rather than the four suggested by Figure 6. The gaps
framework seeks to achieve service component quality by identified by Kyu Kim occur during different stages of the
managing the specification, design and installation of the systems development cycle, such as determining
delivery system for each service component. The frame- information needs, designing and installing the MIS, and
work is shown in Figure 6, and is an adaptation of the delivering the information service (Kyu Kim, 1990).
200 Rands

Service
Technology Environment Needs Component
Past Experience Image

- - - - - -- - - - - -'-~~~~---'

I Technical Gap Service Gap Functional Gap

I
User I Perceived
I Technical F====:~
Perceived Perceived
I Service [Link];::===tFunctional
Service Service
I
I
I
Service
I
' - - - - - - - - - - i Com p onen t 1-- ---'
I Delivery
I ____ Gap 4 ommunication
System
I Gap 1 to
IT I
Gap 3
Users
I
Function I
I
I
I
I Perceptions
of Users'
Ex ectatons

Figure 6 Framework for managing service component quality (source: Zeithaml et al., 1988)

Conclusions system, and the impact of the coordinating mechanisms.

When these, and other research projects, are completed,
In this paper, several service concepts have been it may be possible to evaluate the potential value of these
discussed, and their general meaning in the context oflT aproaches to managing IT. After this new methods of
management examined. At this early stage of applying managing IT may emerge.
service management principles to IT, the discussion in no
way constitutes proof that service concepts are relevant to
managing IT resources and guaranteeing IT effectiveness.
However, their use to explain why the surveys examined
References
by Ives and Olson (1984) were inconclusive is encouraging Bailey and Pearson (1983) Developing a tool for measuring and
for their application to IT. Also, Rahman's (1987) analysing user satisfaction. Management Science, 29, 53~
proposals for building quality into applications through 545.
the software development cycle coincide with the Barki and Hartwick (1989) Rethinking the concept of user
argument by Sasser et at. (1978) that quality cannot be involvement. MIS Quarterly, March, 53-63.
inspected into services. Bowen, Siehl and Schneider (1989) A framework for analysing
In order for the usefulness of service concepts to IT to customer service orientations in manufacturing. Academy of
be evaluated, further theoretical and applied research is Management Review, 14(1), 75-95.
required. Several directions are immediately apparent. Brogonwicz, Delene and Lyth (1990) A synthesises service
Among these is a detailed study of IT service dimensions, quality model with managerial implications. International
possibly identifying dimensions relevant to different IT Journal of Service Industry Management, 1, 27-45.
Bruggeman, Bartholomeesusen and Heene (1988) How
services at various levels of firms. Another avenue is management control systems can affect the performance
research into the value of using SERVQUAL type mea- of service operations. International Journal of Operations
sures for user satisfaction with firms' applications & Production Management, 8, 76-85.
development service. User participation in systems Carman, (1990) Consumer perceptions of service quality: an
development could also be usefully revisited, making assessment of the SERVQUAL dimensions. Journal of
allowances for task dependencies required to develop the Retailing, 66, 33-55.
IT as a service operation 201

Chase, (1981) The customer contact approach to services: Montazemi, (1988) Factors affecting information satisfaction in
theoretical bases and practical extensions. Operations the context of the small business environment. MIS
Research, 29, 698-706. Quarterly, June, 239-256.
DeLone, (1988) Determinants of success for computer usage in Moore (1992) Taking one step beyond quality. Management
small businesses. MIS Quarterly, March, 51-61. Consultancy, May, 67-69.
Doll and Torkzadeh, (1988) The measurement of end-user Parasuraman, Zeithaml and Berry, (1985) A conceptual model of
computing satisfaction. MIS Quarterly, June, 259-274. service quality and its implications for future research.
Galloway & White, (1989) The internal information systems Journal ofMarketing, 49, 41-50.
function as a service operation. International Journal of Parasuraman, Zeithaml and Berry, (1988) SERVQUAL: a
Operations & Production Management, 9,19-27. multiple-item scale for measuring consumer perceptions
Gronroos, (1984) A service quality model and its marketing of service quality. Journal ofRetailing, 64, 12-40.
implications. European Journal ofMarketing, 18, 36-44. Price Waterhouse (1988) Softuiare Quality Standards: The Costs
Harvey, (1990) Operations management in professional service & Benefits (DT!, London).
organisations: a typology. International Journal of Rahman, (1987) Software Quality by Management. Information
Operations & Production Management, 10,5-15. & Softuiare Technology, 29,511-516.
Haywood-Farmer, (1988) A conceptual model of service quality. Rockart, J. (1982) The role of the information systems
International Journal of Operations & Production Manage- executive: a critical success factors perspective. Sloan
ment,8, 19-29. Management Reoieui, 24,3-13.
Hill, (1991) Production/Operations Management: Text and Cases Sasser, Olsen and Wyckoff, (1978) Management of Service
(Prentice Hall, London. Operations Text, Cases and Readings (Allyn and Bacon,
Ives, B. and Olson, M. (1984) User involvement and MIS Boston).
success: a review of research. Management Science, 30, Schmenner, (1986) How can service businesses survive and
586-603. prosper? Sloan Management Reoiet», Spring, 21-32.
Ives, B., Olson, M. and Baroudi, J. (1983) The measurement Shostack, (1982) How to design a service. European Journal of
of user information satisfaction. Communications of the Marketing, 16, 49-63.
ACM, 26, 785-793. Silvestro, Johnson, Fitzgerald and Voss, C. (1990) Quality
Johnson and Morris, (1985) Monitoring and control in service measurement in service industries. International Journal of
operations. InternationalJournal ofOperations & Production Service Industry Management, 1, 54-66.
Management,S, 32-38. Thomson, (1967) Organisations in Action (McGraw-Hill, New
Kierl and Mitchell, (1990) How to measure industrial service York).
quality. Industrial Marketing Digest, 15,35-46. Vickers, Hake and Poore, (1987) Outlook '87 Panel Examines
Kyu Kim, (1990) User information satisfaction: towards Software Quality Issues. Survey ofBusiness, Winter, 13-17.
conceptual clarity. Proceedings of Eleventh International Wild, R. (1989) Production and OperationsManagement (Cassell,
Conference on Information Systems, Copenhagen, December London).
16-19, 183-191. Wild, R. (1980) Operations Management ... A policy Frameuiork
Larsson and Bowen, (1989) Organisation & customer: managing (Pergamon Press, Oxford).
design and coordination of services. Academy of Manage- Zeithaml, Berry and Parasuraman, (1988) Communication and
ment Reoieui, 14, 213-233. control processes in the delivery of service quality. Journal
Lewis, (1988) Customer care in service organisations. Inter- ofMarketing, 52, 35-48.
national Journal of Operations & Production Management,
8,67-75.
Lewis, (1990) Service quality: an investigation of customer care
in major UK organisation. International Journal of Service Biographical notes
Industry Management, 1, 33-44.
Levitt, (1972) Production Line Approach to Service. Harvard Tony Rands is a Fellow in Information Systems and
Business Reoieui, Sep-Oct, 41-52. Management Science at Templeton College, Oxford. He
Martin, R. (1990) The Utilisation and Efficiency of IS graduated from Cambridge in 1962 and was a Systems
(Templeton College, Oxford). Analyst until 1966, when he joined the newly formed
Miller and Doyle, (1987) Measuring the effectiveness of Oxford Centre for Management Studies, later Templeton
computer-based information systems in the financial sector. College. His research interests include developments in
MIS Quarterly, March, 107-124.
the Software Industry and their implications for managing
Mills and Moberg, (1982) Perspectives on the technology of
service operations. Academy of Management Reoieui, 7,
the implementation of IT strategies.
467-478.
Mills and Morris (1986) Clients as 'partial' employees of service Address for correspondence: Tony Rands, Information
organisations: role development in client participation. Systems and Management Science, Templeton College,
Academy ofManagement Reoieui, 11(4), 726-735. Kennington, Oxford OXI 5NY, UK