STRATEGIES FOR CONVERTING TO NEW TECHNOLOGIES
Computer Graphics in the Color Printing Industry
Alan K. Graham
David P. Kreutzer
Sloan School of Management
Massachusetts Institute of Technology
Cambridge, Massachusetts 02139
ABSTRACT
A number of challenges face firms that need to decide when
and whether to convert from traditional technologies to new
computer-based technologies. Such is the case with lithographic
setup shops, which prepare photos for color printing; they must
choose between continuing the traditional craft methods or
acquiring digital image-processing equipment. Pioneering firms
can be saddled with experimental, undependable, and expensive
prototype systems. Rapid technological changes still occurring
in digital systems can allow competitors who invest later to
obtain cheaper, more effective equipment. But firms investing
later may find themselves paying for the large investment just
when most competitors are established in the new technology and
competition has forced prices and profits to low levels.
In order to create an organizing framework for analyzing
and developing conversion strategies for these firms, we worked
in collaboration with Inter/Consult, the project's sponsor, to
build a system dynamics model of the color prepress industry, its
market, and a typical firm. The primary purpose of the model is
to provide a clear understanding of the impact these major
capital investments will have on the profit structure of
lithographic setup shops and to help these shops develop
effective conversion strategies. A secondary purpose of the
model is to aid digital image-processing equipment suppliers in
understanding their market and to provide them with a tool for
generating alternative scenarios given different assumptions
about economic trends, technological developments, prices, market
size and composition.’ The model serves as a strategy support
system that allows clients to derive scenarios explicitly from
causal assumptions and to evaluate alternative investment
strategies.
ee
D-3411 2
STRATEGIES FOR CONVERTING TO NEW TECHNOLOGIES
Alan K. Graham
David P. Kreutzer
Rapid advances in technology are transforming the structure
and the nature of many industries, presenting firms with
difficult challenges. Boundaries between industries change over
time, expanding and shrinking as new applications become
available or firms in other industries capture or replace an
older industry's market. Sometimes separate industries gradually
merge as technologies developed for one industry find
applications in new areas. The traditions, strategies, and
technologies that firms have developed may no longer suffice to
maintain their competitive strength.
Figure 1, based on slides developed by Nicholas Negroponte
and Andy Lippman at MIT, illustrates technological changes in the
computer, publishing, and broadcast industries [1]. Originally,
each technology was entirely seperate; the circles did not
intersect, But the industries have been merging gradually, as
new products are created based on (or applicable to) the
technologies of several industries; so the circles are moving
together. The overlap of computers and broadcast media has
produced computer animation systems and character generators for
print displays on TV. The overlap of publishing and broadcast.
technologies seems ready to produce videotext services, having
already produced information services such as stock market and
credit reporting. When computers and publishing overlap, the
results are word processing and automated typesetting. More
recently, the overlap of computing and printing has produced
digital color processing for printing, which is the subject of
the case study reported here. For a wide variety of businesses
the tools of the trade are changing rapidly, due to diffusion of
new technologies from previously unrelated industries.
D-3411 3
PRINTING
AND WORD PROCESSING
VIDEODISK PUBLISHING
SPEECH SYNTHESIS
aT
DIGITAL
IMAGE PROCESSING
BROADCASTING COMPUTER
AND FILM
INTERACTIVE CABLE
QuBE
SATELLITE A- 1583,
Figure 1. Technological Convergence of Three Industries
Create New Applications (Based on material
developed by Negroponte and Lippman.)
HISTORY OF GRAPHICS IN COMPUTERS
Originally developed for military and space uses, digital
image processing has been just one component of the computer
technologies that have been diffusing into many other industries.
The history of digital image processing and interactive computer
graphics can be traced back to Jay Forrester's Whirlwind Computer
project in the early 1950s, which first used the computer-driven
CRT (cathode-ray tube) display. Another significant development
in interactive computer graphics, which occured during the
12
FACSIMILE TRANSMISSION
D-3411 i
construction of the North American radar net during the SAGE
air-defense program (also directed by Forrester), was the use of
CRTs as command and control processors. Light pens could be used
to touch the image of an airplane or other object on the radar
screen causing the computer to track it as it moved through the
territory. Since then, digital image processing has been
substantially developed by NASA and the Defense Department for
guidance systems for missiles, satellite surveillance, and the
exploration of other planets.
Computers diffused rapidly into the civilian economy.
Because of the intensive computation needed to create and modify
images, however, computer graphics technologies have spread more
slowly. The first such systems were expensive military flight
simulators and specialized applications, such as processing
satellite surveillance data. As computing costs steadily
diminished over the decades, new applications proliferated to
lower-cost systems involving standardized characters for
automated typesetting, word processing, and superimposing text on
TV broadcasts. The increasing power of computers is now making
it possible to compose or manipulate entire images commercially.
Today, computer-generated images are common in film, television,
video games, and advertising, and they are starting to be used by
the subjects of this case study, the companies that prepare color
pictures for magazines and newspapers.
Industry watchers are anticipating rapid growth in
automated offices, videotext, computer-aided design engineering,
and manufacture (CAD, CAE, CAM), computer-aided architecture, and
robotics, all of which can be linked to software and technologies
developed in digitial image processing. In other words, like the
color preparation industry, many industries are about to face the
trauma of conversion to computer-based technologies.
THE COLOR PRINTING INDUSTRY
Similar to the computer industry, the printing industry has
13
D=3411 5
a distinct history with its own brand of technology, starting
with the Gutenberg press in the mid-1400s. Wooden frames and
blocks of movable type were used to press inked images down onto
paper. From those printing blocks developed the highly skilled
craft of engraving and lithography, where artisans would carve
(or etch) their images on metal plates.
During the last century the traditions and descendents of
those craftsmen formed the bases for what is now called a
lithographic setup shop or a color prepress preparation shop.
These shops produce the film that is now used to photoengrave the
metal plates used in modern printing presses. Sometimes they are
a division of a printing company; sometimes they are independent.
Their customers come to them with color photographs and the
outline of a page design, for example, for the color-printed
cover of a magazine or catalog. These printed color pictures are
actually composed of thousands of tiny dots usually in four basic
colors. By juxtaposing these dots precisely, the eyes perceive
the image as containing thousands of colors that subtly and
continously blend together, The setup shops take the original
photographs and text overlays, arrange the page layouts, and
produce separate films of tiny dots for each of the four colors
so that the final product will resemble the original. These
shops also perform image modification such as airbrushing and
color changes. The processes are laborious and demand
considerable skill; so the industry is an ideal opportunity for
computerization.
MERGING OF TECHNOLOGIES
In the past few years computer and printing technologies
have come together so that computer image processing can now
replace many of the traditional methods in the color-printing
industry. Many assumptions about the upcoming changes in this
industry are widely accepted. Digital image processing is
expected to dominate the industry within three to ten years. The
reliability and effectiveness of digital processing is expected
D-3411 6
to increase far beyond that of manual processing as industrial
experience with the technology increases. The cost of the
equipment is expected to drop as the market and technology
mature. Also, as the cost of the equipment drops and its
effectiveness within firms increases, the cost per page job for
the firms with digital equipment is expected to drop sharply.
Given precedents in other new technologies, however,
unstable prices are expected for prepress jobs. When the number
of units in the fiéld nears saturation, the boost in productivity
produces excess capacity, which results in widespread price
cutting by firms attempting to gain market share. This drives
the average market price toward the new, lower cost per job
inherent in digital processing, which reduces the originally
higher profit margin--the very same profit margin that is one of
the main selling points of the equipment!
The expected changes in the marketplace due to conversion
to new technologies pose a challenge to individual firms.
Several strategies exist, each of which presents opportunities
and dangers. Firms must decide what posture to take with respect
to this conversion. Do they want to be pioneers on the
forefront, who perfect the new technology? Should they wait
until after the pioneers have made the major mistakes, determine
the solutions, and then go in? Or should they wait even longer,
until the technology matures, the market settles out, and little
uncertainty remains?
A CASE STUDY: ELECTRONIC GRAPHICS FOR THE COLOR PREPRESS
INDUSTRY
In order to develop an organized framework for analyzing
conversion strategies, we have built a System Dynamics computer
simulation model of the industry, the market, and a firm. The
model examines investment strategies of lithographic prepress
color preparation shops making the transition from traditional
manual processing to electronic color page make-up systems. The
D-3411 7
project was initiated and sponsored by Inter/Consult, a
consulting firm for the graphic arts industry.
SYSTEM DYNAMICS
System dynamics is a methodology and a body of knowledge
about problem solving that revolves around the use of computer
simulation models to test and design policies, much like a wind
tunnel is used to design aircraft. More precisely, Edward B.
Roberts, in Managerial Applications of System Dynamics,
describes system dynamics as "the application of feedback control
systems principles and techniques to managerial, organizational,
and socioeconomic problems [2]." In the seminal text Industrial
Dynamics, Jay Forrester explains how the approach “provides a
single framework for integrating the functional areas of
management-~marketing, production, accounting, research and
development, and capital investment [3]."
Once a computer simulation model is capable of replicating
the historical dynamics of the problem being investigated, a
variety of policy tests can be undertaken, The modeler can do
experiments with the model that would be impossible, expensive,
or risky in real life. "What if" scenarios can be simulated to
compare how well alternative strategies fare under a variety of
possible situations. Finally, new policies, strategies, and
corporate structures can be designed that improve the performance
of the system.
THE ELECTRONIC COLOR SYSTEMS MODEL (ECSM)
The Electronic Color Systems Model (ECSM) simulates the
behavior of the color prepress industry. The primary purpose of
the model is to provide a clearer understanding of the impact
that major capital investments in new processing technology will
have on the profit structure of lithographic setup shops, and to
help these shops develop effective conversion strategies. A
secondary purpose of the model is to aid suppliers of digital
14
D-3411 8
image-processing equipment in understanding their market and
provide them with a tool for generating alternative scenarios
with different assumptions about economic trends, technological
development, prices, market size and market composition.
The Electronic Color Systems Model (ECSM) contains six
sectors. In this paper, only the first four sectors, summarized
in Figure 2, will be active. The first three sectors represent
the changing technological and business environment for the
fourth sector, the firm. The remaining two sectors show
additional dynamics which are of most relevance to producers of
digital equipment, when technological advances enable expansion
into secondary markets with less costly devices. That transition
is in many ways similar to the successive generations of
computers--mainframe, mini, and personal. But space does not
permit discussion of such issues here. Figure 3 is an
overdiagram that shows how the active sectors are interconnected.
Sector Represents
Installed base of electronic color
systems (ECSs) and their
technological effectiveness.
Equipment market
Market size Price per job effects demand for
color setup work.
Market Digital and traditional jobs compete
based on price, delivery delay, and
quality. The firm competes for jobs
based on its relative price, i
delivery delay, and quality.
Firm A typical color setup shop, with
digital capacity and an investment
capability for digital capacity; it
responds to market prices and its
own utilization of capacity in
setting its price. It also
generates indices of costs and
profits.
Figure 2. Summary of Sectors in Electronic Color Systems Model
(ECSM)
15
D-3411 9
/ ,
t
r
i
Pinu | COMPLETED /
PAGE,
ErFicIENcY | EFFICIENCY |
\
\
\
Ants,
THE
FIRM
Figure 3. Structural Overview of ECSM
DYNAMICS OF THE EQUIPMENT MARKET: TECHNOLOGICAL IMPROVEMENT,
THEN SATURATION
Figure 4 shows a policy-structure diagram of the equipment
market sector. This format for diagramming was evolved by
Professor John Morecroft at MIT for use with clients during model
construction [4]. Each circle represents not just one equation,
but all of the equations that define a single policy. This
format is intermediate between a formal flow diagram (which
contains far too much detail for conceptualization discussions
D=3441 10
with non-system dynamicists) and the causal-loop diagram, which
shows none of the levels and rates that are the foundations of
the system's dynamics.
Figure 4 contains only two loops, one positive and one
negative. The positive loop connects effectiveness, installation
rate, and units in the field. It underlies a version of the
classic "learning curve" behavior, where increasing scale of
activity leads to technological improvement, which lowers
costs and raises effectiveness, which in turn leads to a larger
scale of activities.
INDUSTRY'S DIGITAL CAPACITY
Th
Ral
MARKET
SIZE
a,
Figure 4, Policy Structure Diagram of
Equipinent Market Sector
D-3411 a
The model expands the standard learning curve concept. For
electronic color systems (ECSs), functionality as well as cost is
very important. An inexpensive machine that is unreliable and
difficult to understand and use is not a good buy. Also,
technological improvements come from having units out in the
field, so that experience using the units accumulates. This
experience allows flaws in the hardware and software to be
identified and corrected and also permits the designers to
determine which features operators need for the system to be
effective. Finally the costs to the manufacturer come down with
more units in the field, not so much because of accumulated
manufacturing experience, but because R&D costs for software are
spread over a larger number of units. (A substantial amount of
hardware for ECSs is based on off-the-shelf products of standard
computer makers.)
The negative loop in Figure 4 connects units in the field,
potential installations, and the installation rate. The
installation rate is constrained by the finite size of the market
for ECSs. In the extreme, if everyone interested in electronic
color systems had already bought one, then sales and
installations of new units would be zero. (A more elaborate
model might deal with replacement purchases for obsolete
equipment, but such refinements must await an additional project.
This assumption is plausible here because the average lifetime of
the equipment is longer than the duration of the industry
transition being studied.)
Figure 5 shows a simulation of the equipment market sector.
As time progresses, operating experience accumulates, leading to
improvements in the effectiveness of the technology. As
confidence in the technology increases, the installation rate
increases, thereby accelerating the rate of increase in the
effectiveness of the technology. This positive feedback loop
alone would produce exponential growth in the units in the field.
This happens only during the early part of the simulation,
16
D-3411 12
because, as the number of installed units increases, the number
of potential intallations decreases, forming a negative feedback
loop that constrains the growth of the industry. So loop
dominance shifts from the positive loop in the beginning to the
negative loop after the middle, which creates the "S-shaped"
growth of installed units.
INDUSTRY DIGITAL CAPACITY
UNITS IN FIELD (0., 200.)
—— —— INSTALLATION RATE (0.,4.)
——— — — RELATIVE PRICE EFFECTIVENESS OF EQUIPMENT (O.,7.)
FIELD MONTHS OF EXPERIENCE (0., 10.1)
nesilzpemea,
= 7
1.75) Fa SS
? ~~
ate 1-187
90. 120. 150.
‘o TIME
Figure 5. Simulation of Equipment Market Sector
DYNAMICS OF THE COLOR MARKET -~ CONVERSION AND OVERCAPACITY
Figure 6 shows a policy structure diagram of the market
sector, in which the model represents how supply and demand for
color preprocessing influence its price. The market sector also
compares the prices and effectiveness of the firm to the market
averages to determine its order rate for jobs. (This will be
D=3411 3
discussed in the next section.) The market sector can be thought
of as mediating between the equipment market sector (and thus the
installed digital capacity) and the market size sector.
JOB .
REE | pace
| Capacity
9 MARKET
ZE
aH
A-al
MARKET SECTOR
Figure 6. Policy Structure Diagram of Market Sector
Figure 7 shows variables from the market sector, from the
same simulation that generated Figure 5. In the beginning, when
only a few pioneers have electronic color systems, the demand for
color prepress work is roughly equal to the supply, and the price
per digital job is very close to the average market price (even
17
D=3411 14
though the cost per job for digital processing has already
dropped much lower than the cost per traditional job). There is
only slight motivation for those with digital systems to reduce
their prices: the digital price is slightly lower than the
average market price because the trade shops with new
installations are building up their order rates to their new,
higher capacity.
MARKET SECTOR
PRICE PER DIGITAL JOB (O., 8000. )
—— —— COST PER DIGITAL JOB (O., 8000.)
8000. — — — — UTILIZATION OF CAPACITY (O., 1.)
“Tl
ns nee Gene | Fi
6000. \ \ Hl
75 ‘ 5
J \ g
erences ey fi
se
4000. Ss \
)
\
NN NL
2000. = or
225
——
a acius
er) 30, 60. 90. 120. 150.
TIME :
Figure 7. Simulation of Market Sector
By the middle of the simulation, the cost of electronic
color systems is dropping (as modeled in the industry digital
capacity sector) and the number of installations is increasing
exponentially. But when the fraction of jobs done digitally
approaches its maximum, which it reaches by month 80, excess
capacity begins to spread throughout the industry.
D=-3441 15
This initiates pressures causing price cutting and
“shake-outs" that are prevalent in many new industries. Firms
cannot remain cost-competitive without the new technology, so
many remaining firms convert even after excess capacity is
manifesting itself throughout the industry. As more firms
convert, excess capacity gets worse increasing pressures to cut
price, which in turn puts pressure on remaining firms to convert.
This positive feedback loop is exacerbated by the long delays
from the decision to convert, to ordering, delivery, and finally
to bringing new equipment up to full capacity.
It should be noted that in real life there are several ways
for a firm to stay in business while avoiding a direct conversion
to digital technology. Since the color prepress industry relies
on extensive personal contact with publishers, successful firms
that have already converted may either buy out unconverted firms
or take them on as junior partners to act as marketing
organizations for the firm with digital capacity. Also, there
will be room in.the industry for a few traditional firms to
specialize in those jobs which are inappropriate for digital
systems.
Alternatively, firms with excess digital capacity may
accept jobs from other firms at lower than manual cost. The part
of the electronic color system (ECS) that converts photographs to
digital information is much less expensive than the part that
performs image modification. As a result, many firms own the
former (the "scanner") without being able to purchase the latter.
But as scanners become able to transmit the digitized information
over telephone lines, many "retail" shops can send their jobs to
a single "wholesale" firm with a full ECS. Such developments
have already begun in architectural firms converting to
computer-aided design (CAD) systems.
Some words are in order about the technologies and the
industries to which the preceding dynamics apply. Overcapacity
18
D-3411 16
prompts a rapid shakeout in industries where the new technology
is dramatically superior to the old, and that superiority is
directly reflected in the competitiveness of the firm. So
architectural or engineering firms using CAD should show very
similar dynamics. Note, too, that price is not the only
competitive variable; CAD/CAM is catching on very quickly among
automobile manufacturers, not because they gain a significant
price advantage from digital designing, but because CAD/CAM
offers much more timely updating of designs.
By contrast, there seems to be no danger of widespread
overcapacity for word processing users. Most firms do not accrue
any major competitive advantage by converting to electronic word
processing; the conversion decision is a simple cost/benefit
decision with relatively less impact on the overall performance
of most firms.
THE FIRM SECTOR AND POLICY ANALYSIS
The sectors just described provide a background environment
used in testing the consequences of the different strategies an
individual firm uses to make the transition to new equipment in
an industry that is itself changing. Figure 8 shows a
policy-structure diagram of the firm sector, which represents
the policies and conditions within an individual firm: digital
capacity, order backlog, work quality, price and indices of cost
and profit.
D-3411 v7
5 MARKET
SECTOR
FIRU"S ORDER RATE
FIRM.
cost PER
FIRM
Figure 8. Policy Structure Diagram of Firm Sector
The following sections describe generic scenarios generated
by the model. One way the model can contribute to strategy
development is to provide clear pictures of the industry's
behavior patterns the firm is likely to encounter. The model
will only generate behavior that results from a consistent and
explicit set of assumptions about causes and effects, and those
assumptions along with the behavior can be examined for
plausibility--an additional "reality check." (By contrast,
without a mathematical model, it is easy to posit
plausible-looking behavior that could not result from any set of
plausible assumptions.)
19
D-3441 18
The model also offers a method of evaluating different
transition strategies a firm might adopt. For example, the
simulated investment strategies for the firm all take place in
the same market environment. The results within the firm,
however, are distinctly different due to the timing of the
investment decision.
EARLY COMMITTMENT--IMMATURE TECHNOLOGY REDUCES PROFITABILITY
One transition strategy that might be explored by an
aggressive firm wishing to jump ahead of the competition is to
buy the newest, most sophisticated equipment at the first
possible moment. If the equipment is functional, the firm can
realize substantial profits, as it charges traditional prices
while its processing costs are reduced.
On the other hand, if the technology is immature, operating
costs can be enormous. Often, the high initial cost of the
system will not be repayed rapidly because the machine is not
fully utilized. Newly developed systems sometimes stop working
for mysterious reasons, and, surprisingly, sometimes a new system
will sit in the corner unused when a firm is too busy to expend
the manpower to learn to operate it. The problem is compounded
if the system's educational materials haven't been extensively
field-tested and aren't yet effective. Additionally, if a firm
is locked into obsolete equipment, the firms that waited before
buying can put the pioneering firm at a competitive disadvantage
by using cheaper, more effective equipment.
Figure 9 shows a simulation of the early investment
strategy. The firm buys an electronic color system in month 2,
and it is "locked in" to its original level of technology, i.e.,
there is no increase in its effectiveness as the industry gains
experience with digital technology.
D-3411 19
EARLY COMMITMENT RUN
FIRMS PRICE PER JOB (0., 8000.)
——-+—— FIRMS COST PER JOB (0. 8000,
= — — FIRMS UTILIZATION OF CAPACITY (0.,1.>
FIRMS ORDER RATE (0., 400. )
=== Zl
oo 7S g
NE
/ \ 3
/ \
Fi a
7 EH
a / \
a Oa Se
/
/ a
7
Op rssess An 1868
9-0 30. 60. 90. 120, 150.
TIME
Figure 9. Simulation of Early Committment Strategy
A major feature of this simulation, and indeed all of the
strategy evaluations, is the "window of opportunity." For this
firm the window opens at month 45, when digital processing costs
for a new unit drop below the cost of manual processing while the
price charged for processing has not yet dropped. The window
closes around month 92, when widespread overcapacity drives down
prices. With the early investment strategy, the window of
opportunity slams shut particularly hard; since most of the
industry converts to equipment that is more cost-effective than
the firm's equipment, when price-cutting begins, the early-
investment firm will be hit harder than most.
20
D-3411 20
LATE INVESTEMENT--TOO CLOSE TO SHAKEOUT TO RECOUP INVESTMENT
The firm could wait until the technology matures. This
plan avoids the danger of being locked in. The risk, however, is
that the companies which have already converted successfully will
force the traditional firms out of business by reducing their
prices to their new lower costs when excess capacity develops in
the industry. If the traditional firms survive the price
transition and then invest, slender profit margins and thus
erratic cash flows can make financing the equipment both
burdensome and risky.
Figure 10 illustrates the late investment strategy with a
simulation that delays digital acquistion until month 110. This
strategy avoids the early lock-in effect. The equipment
purchased is excellent and embodies a lower processing cost.
However, the firm has also missed the window of opportunity. It
is still paying the extra conversion cost when prices are pushed
to the new cost floor by those firms that have already purchased
advanced systems and payed for them when their profits were
higher. Note that the firm invests when there are still
apparently good profits to be made--the price has just barely
begun to be forced down by the burgeoning excess capacity of the
industry. Such late investment doesn't miss the window of
opportunity due to the immediate profit picture, but because the
price/cost gap will close before the investment can be recouped.
At best, then, the late investment strategy allows the firm to,
stay in business. Depending on the smoothness of its conversion
costs and the competitiveness of the market, it may not even be
able to do that.
D-3411 21
LATE INVESTMENT RUN
FIRMS PRICE PER JOB (0. , 8000.)
—— —— FIRMS COST PER JOB (0. , 8000.)
——— —— — FIRMS UTILIZATION OF CAPACITY (0., 1.) .
FIRMS ORDER RATE (0. , 400.)
|
| Fs
Bi
{ |
3
ra
|
2000. |
+25 t
100.
\ 2
\ 3
(17) soetetatatatatat ietetatettaten iatanaiataien dens. NL
at) 30. 60. 90, 120, 150.
TIME
Figure 10. Simulation of Late Investment Strategy
HITTING THE WINDOW OF OPPORTUNITY -- SUCCUSSFUL CONVERSION
If investing too early locks a firm into obsolete equipment
with high conversion cost, and investing late forces the firm to
endure conversion costs while the industry is undergoing price
cutting there must be a better strategy somewhere inbetween. A
moderately agressive firm whose managers watch the evolution of
the equipment technology closely might take advantage of the
policy of some equipment makers of regularly enhancing the
software and making new, add-on equipment available. Regular
technology updating avoids, to some extent, the problem of older
equipment being locked into early technology. So technologically
sophisticated firms may wish to invest early and acquire
expertise and new customers in hopes that the "updatable"
equipment will remain competitive. Firms investing after the
21
D-3411 22
pioneer stage of the industry are also able to convert more
quickly because the skilled labor and experience of the pioneers
are available.
Figure 11 illustrates a highly successful conversion
strategy by a firm purchasing digital equipment at month 30 just
as the window of opportunity is opening. After recovering from
its conversion costs by month 60 the firm's operating costs
continue to fall until month 110 as it continues to take
advantage of improved efficiency due both to the equipment
supplier's and its own learning curves. This allows the firm to
retain its exceptional profit margin and expand its customer base
even as prices are falling.
SUCCESSFUL CONVERSION RUN
FIRMS PRICE PER JOB (0., 8000.)
—— —— FIRMS COST PER JOB (O., 8000.)
— FIRMS UTILIZATION OF CAPACITY (O., 1.)
FIRMS ORDER RATE (0. , 400.)
1 —
==
FY
z
F
‘i
0 Buv87L
9.0 30. 60. 90. 120. 150.
TIME
Figure 11. Simulation of Succesful Conversion Strategy
D-3411 23
With no additional investment even this successful firm
losses its competiveness by month 140 as it is no longer able to
keep up with the most highly advanced systems available during
the 90 months following its investment,
INCREMENTAL INVESTMENT--HITTING THE WINDOW OF OPPORTUNITY AND
FOLLOWING THROUGH
The firm which successfully invested just as the window of
opportunity was opening up has another opportunity available to
it, That is to follow through with additional system purchases
or major expansions with profits from the first conversion. This
incremental investment approach is perhaps the most attractive of
the investment strategies discussed here. A firm might purchase
an image processor with limited features, then add workstations,
scanners, and sophisticated software packages as its experience
and order rate grow. These major expansions and new systems
improve a firm's capability much more than the enhancements and
upgrades to an already existing system discussed in the previous
section. For this stratagy to be effective one must assume that
newer systems will be compatible with older ones and that
operator skills will be transferable.
Figure 12 illustrates the incremental investment strategy,
in which the firm purchases digital equipment at month 30 and
additional digital equipment once every fifty months after that.
In effect, this is a policy of trying to keep one step ahead of
obsolescence. The three investments have quite different effects
on the firm's competive position. The first investment is the
entree into electronic color systems; the investment provides
experience and begins to build order volume. Profits are
substantially improved, until improvement in ECS technology and
price cutting begin to threaten profits. The second investment
remedies the situation by assuring that the firm's cost after the
transition will be as low or lower than those of its competitors.
The third investment does not provide additional technological or
22
D-3411 24
cost advantage, but does continue to expand the size of the
firm--revenues expand even after an era of price-cutting.
INCREMENTAL RUN
—————FIRMS PRICE PER JOB (0., 8000.)
—— —— FIRMS COST PER JOB (0,, 8000. >
——- — FIRMS UTILIZATION OF CAPACITY (0., 1.)
FIRMS ORDER RATE (0., 400.)
A H
6000. ‘\
300. 71
AS 1812.
‘9.0 30. 60. 0. 120. 150.
TIME
i
Figure 12. Simulation of Incremental Investment Strategy
By constantly increasing its own effectiveness, the firm
competes with increasing effectiveness and decreasing price
industrywide. The run shows that the firm's constant price
cutting leads to ever-increasing orders.
There are a number of ways a real firm can attract business
as its capacity expands. One, of course, is to attract customers
one at a time with more favorable pricing and superior service.
D-3411 25
Others involve taking over unconverted firms and their customer
bases either through outright ownership or through various forms
of subcontracting to the unconverted firms, which in effect
become purely marketing organizations.
Naturally, there are pitfalls to continual expansion. For
example, development technology for photographs advanced to the
point where chains of specialized shops using small developers
could promise overnight developing. One of the more prominent
chains went bankrupt when, after taking an unexpectedly long time
to build up a customer base, its cash commitments for expansion
exceeded its revenues, which were already depressed by the state
of the economy.
All in all, the scenarios analyzed here indicate that the
best strategy is to convert during the window of opportunity.
This result may seem trivial at first glance, in that the best
strategy is a compromise between two extremes. But the
triviality is more illusion than real. The model provides fairly
specific characterizations of what to look at to determine where
the industry is relative to the window of opportunity. The
opening of the window occurs when only a few other firms are
fully utilizing the new equipment. The window for profitable
investment is already closing when a substantial fraction of
firms in the industry have converted--this is before a shakeout
occurs. And even firms that failed to convert until after the
shakeout has started will be given considerably more impetus by
the model to make realistic plans.
CONCLUSION
As the work on strategy evaluation evolved, the "window of
opportunity" concept became more and more central. Also it
became clear, as the remarks from time to time indicate, that the
results should apply to many industries. Before finishing the
substantive results, then, it would be well to ask "under which
circumstances will there be a window of opportunity to take
23
D-3411 26
advantage of?"
Figure 13 summarizes four prerequisites for the existence
of a window of opportunity. First, as previously mentioned, the
technology must have a major impact on the firm's competitive
performance, otherwise there will be no shakeout. Second, the
technology must have prospects for major improvement. This must
be taken relatively, for everything is improvable to some degree,
but if there is no need to worry about large improvements in
later models, one can convert to a new technology whenever it
appears cost-effective. For example, few of us would postpone
buying a car for next year's models--we simply don't expect
enough improvement to outweigh the inconvenience of postponement.
Technology has major impact on firm's competitiveness
Technology has prospects for substantial improvement
Technology expands scale of efficient production
Intermediate speed of penetration
Figure 13. Prerequisites for the Window of Opportunity
Third, the technology must expand the scale of
production. This is actually a requirement not only of the
technology, but of labor and business practices: converting to a
new technology must increase the amount of production desired by
firms (otherwise conversions will reduce the number of employees
rather than produce overcapacity). When AT&T converted to
electromechanical switching during the 1930s and 1940s, the
conversion did not produce overcapacity; the switchboard
operators were gradually discharged. But for lithographic setup
shops, once an electronic color system is in place, strong
pressures exist for individual firms to try to utilize it fully
by expanding operations, And a number of factors make it
difficult to reduce "manual capacity"--i.e., skilled artisan
employees. Union pressure, personal ties, and uncertainty about
the new technology all motivate management to retain employees
D-3411 27
and attempt to expand the business. When the entire industry
retains both employees and digital capacity, the conditions for a
shakeout are created. (This effect can be mitigated to some
extent if the drop in prices cause a large increase in the number
of color jobs due to conversion from customers for black and
white jobs. The model has structures to simulate and test these
dynamics, however, they were turned off for the runs in this
paper.)
Finally, technology must not penetrate its user market
very rapidly (two months) or very slowly. If an industry adapts
a new technology all at once, the window of opportunity is gone,
but then so is the question of when to convert. By analogy, no
child obtained any advantage from being a trend-setter with
hula-hoops, because everyone had one. At the other end of the
spectrum, if so few firms convert that the technology never takes
off (even if it is ultimately improvable), there is no shakeout
and no window of opportunity.
Turning away now from substantive results, the project had
some interesting methodological results as well. This study was
among the first applications to use Morecroft's policy-structure
diagrams extensively for model conceptualization, client
interaction, and research organization [4]. Even though the
consultants with whom we worked had no formal training in System
Dynamics notations, the diagrams provided a very efficient focus
for discussions of both the subject matter and the progress of
the project. Moreover, in a good example of technological
synergy, the diagrams were implemented on a CAD/CAM system, so
they were easily kept both up-to-date and neat from the very
beginning of the project [5].
Unlike many of the classic system dynamics studies, the
model's primary use was to generate a plausible scenario to aid
in understanding the system. Using a dynamic simulation model to
generate scenarios seems to offer some advantages over the
24
D-3411 28
"mental generation" of static scenarios practiced in some
prominent think-tanks: The model scenarios follow demonstrably
from the present time. By contrast, with a static,
mentally-generated scenario, one can have little confidence that
there exists any plausible sequence of events that would bring
the scenario into existence, A trivial example: One can easily
imagine a world in which everyone is superbly educated, and
productive enough so that no one starves. While this is an
inspiring scenario, its most important aspect is missing--how do
we get there from here?
The research reported here was sponsored by a consulting
company, Inter/Consult, that wanted to be able in turn to use the
model with their clients. Those clients need a way to gain
confidence in the model without having to build it, or to
understand System Dynamics concepts. Responding to this need, a
follow-up to this project is developing a user interface to allow
clients to examine or modify parameters in the model and simulate
it with minimal guidance [6]. The clients seem to appreciate the
ability to adjust the model toward having the same numerical
characteristics as their own operations.
Although the modeling effort generated the several
interesting methodological results, it also achieved its
principal goal, which was to improve the understanding by various
clients of the issues surrounding the decision to convert to new
technologies. Although project results like "improvement of
understanding" are certinly difficult to assess, the authors!
subjective impressions are that the study contributed an enormous
amount of clarity and structure to a situation that was
previously little more than a loosely related tangle of worries
about the future. Indeed, those involved in the project are
beginning to exhibit the "that's so obvious that we must have
always known it" syndrome that too often hides the true utility
of system dynamics studies.
D-3411 2
ACKNOWLEDGEMENTS
The modeling project whose results are described here grew
out of a class project under the auspices of the MIT Corporate
Research Program in System Dynamics [7], which was created by
Prof. John Morecroft [8][9]. The work was supported both
financially and epistomologically by David Goodstein and Gregory
Van Buren of Inter/Consult. Robert Lucadello performed part of
the original modeling work. The policy-structure diagrams were
executed by David Kreutzer and Janet Gould with assistance from
Scott Ewing on a Designer System donated to M.I.T. by
Computervision Corporation and made available for use by
Professor David Gossard [5]. Diane Senge drew Figure 1. Andrew
Plummer assisted with preparation and editing.
REFERENCES:
[1] Lippman, Andrew B. and Negroponte, Nicholas, Slides from the
Arts and Media Technology Case Studies, 1983, Architecture
Machine Group, Massachusetts Institute of Technology.
[2] Roberts, Edward, Managerial Applications of System Dynamics,
1981, Cambridge, Massachusetts: The M.I.T. Press, Page 3
[3] Forrester, Jay W. Industrial Dynamics, 1961. Cambridge
Massachusetts: The MIT Press.
{4] Morecroft, John D. W., "A Critical Review of Diagramming
Tools for Conceptualizing Feedback Systems Models." System
Dynamics Group Working Paper D-3249-3, Sloan School of
Management, M.I.T., Cambridge, Massachusetts.
{5] Gould, Janet and Kreutzer, David, "Interactive Computer
Graphics Technologies for Representing System Dynamics
Model Structure, Behavior, and Results," System Dynamics
25
D-3411 30
Group Working Paper D-3431, Sloan School of Management,
M.I.T., Cambridge Massachusetts.
[6] Van Buren, Greg and Goodstein, David H., "Development of a
Causal User Interface for Simulation Models," 1983,
Inter/Consult, Technology Center, 21 Notre Dame Ave.,
Cambridge, Massachusetts.
[7] Kreutzer, David P. and Lucadello, Robert, "Strategies for
Investing in Electronic Color," 1982, System Dynamics Group
Working Paper D-3357, Sloan School of Management, M.I.T.,
Cambridge, Massachusetts.
[8] Morecroft, John D. W., "The System Dynamics Corporate
Research Program: A Prospectus," 1983, System Dynamics
Group Working Paper D-3245-9, Sloan School of Management,
M.I.T., Cambridge, Massacusetts.
[9] Morecroft, John D. W., "Administrative Science and System
Dynamics: Filling a Gap in Management Education," 1983,
Sloan School of Management Working Paper WP-1429-83,
M.I.T., Cambridge, Massachusetts.
CC BY-NC-SA 4.0