Michael D Zisman.

Office automation : revolution or evolution? online

. (page 1 of 2)
Online LibraryMichael D ZismanOffice automation : revolution or evolution? → online text (page 1 of 2)
Font size
QR-code for this ebook


LIBRARY

OF THE

MASSACHUSETTS INSTITUTE

OF TECHNOLOGY





Center for Information Systenns Research

Massachusetts Institute of Technology

Alfred P Sloan School of Management

50 Memorial Drive

Cambridge, Massachusetts, 02139

617 253-1000



Office Automation: Revolution or
Evolution?

Michael D. Zisman



CISR #34
Sloan WP 986-78

April 1978



Note: A slightly revised version of this paper will
appear in the June 1978 issue of The Sloan
Management Review .



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 1

INTRODUCTION



1.0 INTRODUCTION

One of the great shortcomings of the information
systems industry has been our inability to manage
expectations. This is not surprising when one considers the
frequency with which we announce a new management concept or
technological breakthrough which will inevitably
"revolutionize" the means by which we manage our information
resources. In the past ten years, we have thrust upon the
world the "revolutions" of timesharing, database management,
the minicomputer, the "total" MIS, distributed data
processing, distributed databases, structured programming,
microcomputers, and now upon us is the "revolution" of
office automation. Each of these proclaimed revolutions has
led the user community to a level of expectation which is to
a large extent unattainable, at least in the timeframes
predicted.

Recent literature has suggested that there will soon be
a revolution in the office as computer technology is brought
to bear to increase the productivity of our white collar
labor force. The "widespread" use of word processing and
the introduction of electronic mail systems are cited as
evidence of the oncoming revolution of office automation.
Although we are led to believe that office automation is a
relatively new phenomenon, office automation is rather an
old term given new meaning. In the late sixties, the term



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 2

INTRODUCTION



office automation referred to the application of computers
to well structured, high volume office tasks such as payroll
processing, accounts payable, purchasing, etc. In its
present reincarnation, "office automation" refers to the
application of computer technology, communications
technology, systems science and behavioral science to the
vast majority of less structured office functions which have
not been amenable to traditional data processing technology.
In both cases, however, the objective of office automation
is to improve the productivity of our white collar labor
force.

While there can be little doubt that technology has
provided an opportunity to dramatically change the ways in
which we perform administrative processing, the rate at
which this technology will be integrated into most
organizations will not be revolutionary at all. Rather, a
scenario in which office environments evolve and adapt to
this new technology is a much more reasonable one. The
purpose of this paper is to address the evolution of office
automation technology. Although this approach may seem less
dramatic, we suggest that it is far more realizable and will
provide managers a framework with which to realistically
plan for the integration of new technology into the office.
We will view this evolution in the framework of Nolan's
stage hypothesis for the development of information systems



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 3

INTRODUCTION



organizations [6], and use this framework to distinguish
"office automation" from "office mechanization". This will
be followed by a discussion of some recent research results
which will impact and provide direction for this evolution.

Before delving into an examination of Nolan's stage

hypothesis, it will be useful to motivate the discussion by

considering why the office is now "ripe" for further
automation.



2.0 WHY OFFICE AUTOMATION NOW??

Why are we interested in office automation? What
benefits are to be gained by applying this computer
technology to millions of office workers who make up our
white collar labor force? Some statistics shed liRht on
these Questions.

Traditionally, the organization has made little
investment in its office employees. Estimates for average
capitalization per office worker vary from $2000 [^] to
about $6000 [16]. By comparison, the average capitalization
per factory worker is reported to be about $25,000 [4].
Organizations have made fundamentally different
capital/labor decisions in these two environments. In the
factory, opportunities for automation have led the
organization to trade labor for capital investment; this



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 4

WHY OFFICE AUTOMATION NOW??



has resulted in impressive productivity increases. In the
office, however, there has been little interest in
automation opportunities.

Current cost trends now make automation in the office
more attractive. Presently, labor costs in the office are
increasing at 6% per year, while communications costs are
declining at 11% per year, computer logic costs are dropping
at 25% per year, and computer memory costs are plummeting at
a rate of 40% per year [33. When one considers that
approximately 22% of the U.S. labor force is involved in
"office work" or "administrative processing" and that the
shift toward a more service based economy will likely
increase this factor, the implicationss are very clear,
indeed.

Until recently, offices have taken little advantage of
technological opportunities to increase productivity. This
has resulted in office costs increasing from 20% - 30% of
total company costs to 40% - 50% of total costs [4].
Furthermore, accounting for these costs is very haphazard in
most organizations, where a large portion of these costs are
lumped into corporate overhead [16]. The decreasing cost of
office automation technology, as well as the availability of
that technology, coupled with the increased cost of office
labor, provides an opportunity for innovative corporate
managers to stabilize or reduce office costs.



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 5

WHY OFFICE AUTOMATION NOW??



Many organizations now realize that they must begin to
move toward the automated office if they expect to stem the
rising tide of office costs and still process the growing
workloads being placed on them. A Stanford Research
Institute study suggests that average investment per office
employee will rise from $2000 to $10000 by 1985 [4]. If
this is the case, the total market for office systems
technology could reach $85 billion by the 1985-1990
timeframe.

On the negative side, the availability of sophisticated

technology at relatively low prices has led some individuals

to believe that technology alone will solve the problems of
the office. This is a delusion. Technology alone provides

opportunities. If the technology is to be of use to an

organization, it must be incorporated into a systems
solution to the problems of the office.



3.0 STAGES OF GROWTH

A stage hypothesis proposes that a specific
evolutionary process can be conveniently segmented into a
number of clearly identifiable phases for purposes of
discussion and comparison. A good example is the division
of the human growth process into stages of infancy,
childhood, adolescence, adulthood and old age. Obviously,



OFFICE AUTOMATION: REVOLUTION OH EVOLUTION? Page 6

STAGES OF GROWTH



any one of these stages could be further subdivided if needs
so dictated. As such, a stage hypothesis is useful to the
extent that it provides insights into the evolutionary
process under study.

Nolan suggests that there exist four stages of growth
for an organization as its use of information systems
matures. He labels these stages initiation, expansion,
formalization, and maturity. These stages lead to the
familiar "S" growth curve. The office will soon become a
major applications area for information systems
organizations and we foresee the office undergoing similar
stages of growth. We examine these stages in more detail so
that managers can place present and future developments into
a broader perspective. This will be followed by an
examination of other relevant hypotheses.



3.1 Initiation

In the initiation stage, organizations will perceive
(or have perceived) technological opportunities for
cost-reduction or increased productivity and will begin to
use mechanized office equipment. This will normally be
limited to text processing equipment, just as the first
computer installations were limited to bread and butter
applications. Some users will take advantage of the machine



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 7

STAGES OF GROWTH



readable text and will interface to photocomposition
systems, TWX/Telex and other output mediums for enhanced
output capabilities. Many of today's larger organizations
are already in or through the initiation stage, using
standalone word processing equipment, shared logic systems,
or text editing systems on large computers. Emphasis will
be on the more efficient production of paper (as opposed to
the longer range objective of reducing paperwork), and in
most organizations the technology will be introduced and
managed by the administrative services function. There will
be a fairly clear delineation between word processing and
data processing at this point. Unfortunately, most data
processing managers will view this stage with little or no
interest, viewing word processing as an activity outside the
scope of their mission.



3.2 Expansion

The initiation stage will be followed by a stage of
rapid expansion. Office automation will "catch on", just as
data processing did during the second stage of EDP growth.
There will be an explosion of interest and predictions of
revolution; poorly planned and uncoordinated development
efforts will be justified on low marginal cost, given the
equipment in place from stage 1 .



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 8

STAGES OF GROWTH



The emphasis in this stage will be on the development
of tools by mechanizing devices in the office such as the
in-basket, file cabinet, Roladex, calendar, telephone, etc.
Most references to "office automation" or the "office of the
future" really refer to this stage of development, and would
be mopre accurately labelled "office mechanization". The
objective here will be to replace paper flow with electronic
information flows. We note that the technology, to a large
extent, exists- today to develop sophisticated document
generators, computer based message systems, basic calendar
management functions, and other related office tools. In
this stage, the data processing sector of the organization
may recognize the office as an applications area.

We will mechanize tasks that people perform (i.e.,
typing, filing) but not automate functions (e.g., order
entry, credit approval). Here we are not faced with a
technology problem as much as a systems problem and an
organization problem. This stage will bring with it the
introduction of large amounts of distributed hardware. For
many of these applications to be successful, they must be
used by a "critical mass" of users (electronic mail is a
good example of this). As more and more people begin to use
these systems, we will have to address the organizational
problems which will undoubtedly arise. The challenges of
the organizational problems will probably be larger than the



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 9

STAGES OF GROWTH



technical challenges in this stage. The proliferation of
applications and hardware will be very similar to the
proliferation of applications in the corporate computer
facilities in the 1960's.

Since the emphasis in this stage is on the
mechanization of tools in the office as opposed to office
functions, per se, there will not be great concern for the
integration of these tools into a cohesive whole. Rather,
it will be a period of experimentation as tools are
proposed, developed, refined, and some discarded.



3.3 Formalization

The period of rapid expansion will result in a
proliferation of disjointed tools, very analogous to the
proliferation of non-integrated data processing applications
of the 1960's. As users become more comfortable with these
tools, they will demand cohesiveness and further integration
in their office environments. This will motivate a third
stage of growth which will focus on what is being done in
the office in addition to the tools used in the office to
carry out office functions. The first change will be toward
integrating applications and facilities into more cohesive
systems. The second change will be a shift from mechanizing
devices toward automating processes, and from mechanizing



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 10

STAGES OF GROWTH



ta sks to automat ing funct ions. A substantial portion of the
work carried on in most offices is the execution of routine
or "almost routine" office procedures. It is these
procedures which give utility to the various devices in the
office and the automation of these routine and almost
routine procedures will allow us to use the computer in the
office as more than a "word cruncher". The processes that
we are addressing here are those office procedures which are
not amenable to conventional data processing technology due
to a lack of structure.

This shift in emphasis from devices to process is a
crucial one. Ackoff [1] provides a succinct definition of
process as a "sequence of behavior that constitutes a system
and has a goal producing function." It is the manager in an
organization that has a notion of process. From a very
simplistic viewpoint, the devices in the office are
resources of the secretary, and the secretary is a resource
of the manager. By mechanizing devices, we are addressing
the resources of the secretary but, by and large, not those
of the manager. However, when we attack office processes,
we are automating office functions, not office tasks, and
are addressing problems of the manager. We are not
suggesting that secretaries will be replaced; we do suggest
that some of the more structured, routine and mundane
responsibilities of the secretary and manager will be



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 11

STAGES OF GROWTH



(almost) completely automated in this stage, thus making
this group of workers available for more productive
functions. Naturally, the difference between stage 2 and
stage 3 is one of degree. In stage 2, we mechanize many of
the secretary's tools but still rely on the secretary to
know when these tools are to be used and how they are to be
used to accomplish some job function. In stage 3, we take
the organization's knowledge about the office processes and
put that into the machine so that the computer system will
know how the various tools should be used together (the
sequence of behavior) to accomplish some particular job
function (the goal producing function), and when to use the
various tools. In this stage, we begin to use "automation"
in the stricter sense of replacing people with machines,
rather than simply supporting or augmenting them with more
mechanized devices. This focus on office automation goes
bfeyond many current perceptions of the field.



3.M Maturity

The last stage in the growth process is a relatively
self-explanatory one, and a period of stabilization as the
organization adapts to change. We will see a continued
integration of functions and facilities in office systems.
Less structured office procedures will be automated as we
learn more about this area. Obviously, some functions will



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 12

STAGES OF GROWTH



be "de-automated" as we learn from the mistakes. As major
new technologies develop, this growth curve will be
repeated, as organizations generate change, integrate the
changes into their work environments (possibly by changing
the work environment), and stabilize from the effects of
these changes.

3.5 Other Stage Hypotheses

Strassmann [17], of Xerox, has also considered the
implications of the stage hypothesis on office automation.
He suggests that while traditional data processing may be in
the fourth stage of Nolan's growth curve, "general
administrative systems" are just beginning to emerge,
largely because they have been ignored for the past ten to
twenty years. Strassmann sees the four stages of growth in
general administrative systems to be mechanization of tasks,
machine-aided transactions, work redesign and work
enlargement. The work reported in this paper concentrates
on what Strassmann has labelled task mechanization and
machine-aided transactions, and is complementary to
Strassmann's model.

In the late sixties, a popular stage hypothesis [15]
for office automation included the nonmechanized craft
stage, an early mechanization stage characterized by the use



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 13

STAGES OF GROWTH



of typewriters, adding machines and the like, the punched
card data processing stage and the automation stage of
electronic data processing. Clearly the stage hypothesis
discussed in this paper, and that discussed by Strassmann,
deals solely with what was previously considered the
automation stage of electronic data processing and explodes
that single stage into mechanization and automation
components. These views are not necessarily inconsistent if
one views the distinction between mechanization and
automation as a technology frontier which is continually
moving forward; what we view as automation today will
probably be viewed as mechanization is some future period.

M.O IMPLICATIONS OF THE STAGE HYPOTHESIS

Our view of the evolution of office automation clearly
revolves around a distinction between mechanization of tasks
in stage 2 and automation of functions in stage 3. We
believe that there are several implications for those
charged with managing this evolution, and we discuss some of
those implications in this section. Before proceeding to
that, however, it will be necessary to more clearly
distinguish the terms mechanization and automation.



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 14

IMPLICATIONS OF THE STAGE HYPOTHESIS



M.I Mechanization And Automation

It is distressing that most of the current literature
on office automation addresses the oncoming "revolution" in
the office but makes no attempt to distinguish office
automation from office mechanization, and in fact, often
mistakes one for the other. Earlier literature on the
subject of office automation studied the distinction between
mechanization and automation. In addressing this
distinction, Shepard [15] provides a good summarization of
the literature. Two major points distinguishing
mechanization from automation according to studies by
Faunce, Bright and Diebold are:

1. As automaticity increases, the initi a ting control
source moves from man to technology (i.e.,
automatic control);

2. As automaticity increases, integration of function
increases.



As pointed out in the previous section, it is these
characteristics which distinguish stage 2 from stage 3. The
second point, that of integration of function, is fairly
self-explanatory. The first point, that of control, is
crucial to real automation. In the office domain, this
means that we not only mechanize the task (e.g., from



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 15

IMPLICATIONS OF THE STAGE HYPOTHESIS



typewriter to word processor) , but that we also use the
technology to know when a task should oe executed (e.g.,
under what conditions the word processor should be
instructed to produce a specific document) . As Simon
indicates [14J , technology is knowledge; "knowledge of how
to make things, but also knowledge of how to do things." The
technology of automation implies a different kind of
knowledge and a higher degree of knowledge, than the
technology of mecnanization. In the office domain, this
incremental knowledge is, in fact, the knowledge of office
procedures.

4.1.1 Automating Office Procedures -

In order to automate office procedures, we need a
general mechanism for representing the knowledge about
office procedures in the computer memory. In this way, the
computer can automatically react to inputs to the office by
activating the various mecnanized tools with instructions
supplied by the encoded procedures. This issue of office
process representation is an instance of a more general
process representation and knowledge representation problem
which is of wide interest in the artificial intelligence and
computer science research communities. What we are really
addressing is the development of knowld ege based systems .



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION?
IMPLICATIONS OF THE STAGE HYPOTHESIS



Page 16



In considering the representation and automation of
office procedures, certain characteristics become apparent
which can be exploited. Foremost is the extent to which
office procedures are event driven. By that, we mean that
many office procedures are defined in such a way that office
personnel react to various stimuli from the environment. A
reasonable paradigm for an office manager is to "keep the
ball in someone else's court." When something happens which
requires our attention, we react to it, and then send it on
to the next station for further action.



As part of a research project to develop a suitable
representation for office processes, the author worked with
a large mortgage banking company and large insurance company
in analyzing their office procedures. To a surprisingly
large extent, their procedures were event driven and could
be characterized by what we came to call a "local focus of
attention." The procedure is composed of a number of related
activities. An activity is an action or series of actions
which occur in a single attention span, is frequently
performed by a single individual, and is initiated by some
external event (external to the person performing the
activity) such as the arrival of a document or the arrival
of some specific date. An activity terminates when no other
actions can be taken with respect to the procedure.
Normally, this means that the organization must await new



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION?
IMPLICATIONS OF THE STAGE HYPOTHESIS



Page 17



information from the environment or must wait for some fixed
time period to elapse.

Another interesting aspect of office procedures is the
observation that oftentimes difficulties in offices arise
not so much from doing the wrong things as from not doing
the right things when they should be done. The difficulty
is not task performance as much as task recognition. This
is an area in which the computer may be very helpful.



4.2 Organizational Implications

Much of the work cited earlier which distinguished
mechanization from automation was actually motivated by
questions of worker alienation in offices and factories. In
general, the literature indicates that workers in automated
environments are less alienated than workers in mechanized
environments. This is generally due to job enlargement and
less function specialization in automated environnments
[15]. Shepard shows that for both the office and the
factory "the degree of differentiation in the division of
labor at work is related to technology and that automated
technology reduces the level of alienation among office
employees as well as factory workers."



OFFICE AUTOMATION: REVOLUTION OR EVOLUTION? Page 18

IMPLICATIONS OF THE STAGE HYPOTHESIS



An obvious example is the factory production line. In
a mechanized production line, each worker makes a small
contribution to the final product as the assembly moves by
each work station. The job is characterized by mechanical
pacing of work, repetitiveness, minimum skill requirements,
surface mental attention and minute subdivision of labor
[18]. These workers are likely to have relatively high
levels of alienation. In contrast, the worker in an
automated continuous process plant is a monitor of an
integrated production system, a maintainer of complex
equipment and an exception handler. These jobs are
associated with the whole of the process, as compared to the
high subdivision of labor associated with mechanization, and
job enlargement usually occurs.

We can expect the same results in the continuing
evolution of office systems technology. During periods of
device mechanization we can expect to encounter substantial
employee resistance and alienation. Large word processing
centers provide a good example of a highly mechanized,


1

Online LibraryMichael D ZismanOffice automation : revolution or evolution? → online text (page 1 of 2)