Organization Design and the Dynamics of Performance Over the Long Term
Henry M. Frechette, Jr
Senior Vice President
The Forum Corporation
Frank Spital
Associate Professor
College of Business Administration
Northeastern University
Abstract
This paper addresses the issue of organizational design and its impact on long term business
performance. There is evidence that organizations progress through a series of cycles, but
the organization design literature typically does not consider these cycles. We examine the
organization design recommendations of Forrester in terms of their impact on delays in a
company. We then mn a small simulation to explore the impact of these delays on long
term company performance. We conclude that most organization structures act to impede
long term performance.
The Problem
The average life of companies is unexpectedly short. Even substantial companies fail at an
alarming rate (Davis and Davidson, 1991). Years ago Greiner (1972) suggested that
organizations go through predictable phases of evolution, when there is no major upheaval,
interspersed with bursts of revolution when there is intense change. He further argued that
the sequence of evolution and revolution is predictable and most organizational failures
occur during periods of "revolution". Organizations which successfully navigate through
periods of revolution do so by instituting new practices during these periods. Managers
should not act to avoid these phases because each phase results in certain strengths and
learning experiences essential for success in subsequent phases. Revolutions provide the
pressure, ideas, and awareness that afford a platform for change and the receptiveness for
the introduction of new practices. But policies that had contributed to solutions in one phase
also become the reason for problems in subsequent phases. In short, the practices and
structures which lead to success at one time become the limits-to an organization’s growth
at a later time.
This paper explores the implications of system based managerial and organizational design
principles on these performance cycles. We argue that the core of the problem is managing
the delays of information flowing through the organization. These delays in turn affect the
long term performance of the organization.
We agree with Greiner and others who have suggested that cycles are not only unavoidable
but may be beneficial to the long term survival of the organization. Besides killing
companies, the periods of revolution, or strategic change, provide the impetus for the next
period of evolution. They have the potential for re-inventing the organization.
- 183 -
The issue is managing these cycles of prolonged stability followed by major upheaval.
Forrester (1965) and others have proposed alternative organizational design and
management structures that they expect will lead to long term success. Forrester, in
particular, makes some interesting suggestions. We examine these ideas in the context of
their effect upon these cycles. Primarily, we will consider how Forrester’s recommended
policies affect delays.in the organization, and it’s response to the need for change.
Forrester’s (1965) work recommends changes to six components of organizational design if
an organization is going to achieve its potential. We suggest that a seventh component
needs to be included—a common organizational purpose.
For each of the design elements to have the results intended by Forrester, they must be
operating in the context of a clearly articulated and understood organizational purpose.
Without this, each of the recommendations may hinder performance by increasing
fragmentation of the organization and reducing the clarity of focus or objectives. These in
turn add delays to the system. Briefly, we will now review each of Forrester’s main points
within the context of a purpose driven organization.
Eliminating the superior-subordinate relationship, particularly multiple layers of such a
hierarchy, should decrease the lag between an action taken by an individual and its feedback
effect from the market upon that individual. With a hierarchy, the feedback is from the
superior or from the control mechanism (budget and planning system), and this feedback,
even if it reflects the market, will have substantial delays. More likely the feedback will
reflect many issues in addition to the market.
Forrester suggests replacing the superior subordinate.relationship as it currently exists in
most organizations with individual profit centers. His argument is that ultimately this is the
fairest means for establishing individual worth. It is not clear that individual profit centers
affect information delays in any predictable way. Unless people are being influenced by the
overall purpose of the organization, and feedback from the end customer, the gains from
reduced hierarchical delay may be negated by negotiations based on self interest rather than
organizational purpose.
Forrester further argues that policy is most damaging when it completely defines action and
states exactly what is to be done (is low in freedom), is undeterminable and subject to future
definition and retroactive determination (low in accessibility), is internally inconsistent and
provides no guide for resolving these conflicting pressures (low in consistency), and is
imposed on a subordinate for the benefit of the superior. The first set of conditions
suppress innovation and initiative, and increase antagonism and confusion. These should,
in turn, delay the ability of the organization to respond to the need for change.
Forrester argues that information in. organizations is late, incomplete and distorted. He
suggests the introduction of a corporate EDP system in "hub and spoke" architecture. We
believe that such a system will increase the speed at which information moves, though not
necessarily its completeness or accuracy. The latter will be primarily determined by the
clarity of the organizational purpose. To the extent that the design of the system reflects
the purpose of the organization, it will reduce the delays in the organization.
Another of Forrester’s recommendations is freedom of internal movement. We believe this
will help provide an integrated and faster response to changing environmental conditions.
We believe the same argument can be made for education; and particularly for just-in-time
education. This can reduce delays in the system by clarifying the participants understanding
of the conditions that they are in, and of the likely solutions to those conditions. This is
based on the assumption that the training and education is supporting the requirements of
the whole system, not simply part of it.
Simulation of Delay Effects
We have identified the impacts that the above design characteristics will have on speed of
response of a company. For the most part, we believe that the design characteristics
recommended by Forrester, when combined with a shared purpose, will reduce the delays
in information flow within an organization. We now turn our attention to the effects of
these delays on organizational performance. To explore these issues we have run a small
simulation. The causal loop diagram of the model is below. (For more detail on this model
see Frechette and Spital, 1991).
The results of putting delays into this model are shown in graphs 1-4. The first two graphs
illustrate the effects on long term performance of increasing delays within the incremental
change loop. These delays are between perceived performance gap and readiness for
incremental change (graph 1), and between readiness for incremental change and start
incremental changes (graph 2). The second two graphs illustrate the effects of delays within
the strategic change loop. Graph 3 is the model output with delay between perceived
performance gap and readiness for strategic change. Graph 4 is the model output with
delay between readiness for strategic change and start strategic change. Our performance
measure is cumulative performance gap, where the gap is the difference between expected
business performance and actual business performance. Therefore a smaller, or even
negative performance gap indicates improved organizational performance. We report the
cumulation of this gap in order to make the longer term outcomes clearer.
- 185 -
Graph 1, Delay before "readiness for incremental change" Graph 2. Delay before "start incremental changes”
aan
243.71
RA PRR
een Ney ;
ann. rae. SY
MN:
“\ . WA
W_
2200.4 5
00 todas 200.50 1300.75 wor ry 08.25 700.50 300.75 01.00
vars Gears
Graph 3. Delay before “readiness for strategic change” Graph 4. Delay before “start strategic change”
/
iS xa
5 a Am - 1605.1
ers 08.25 rola Tobe 40100 00 106.25 geese
varers
Graphs 1-4
Run # Delay ;
1 0.00 Output variable:
2 8.00
3 16.00 Cumulative performance gap
The results shown in graphs 1-2 indicate that an eight period delay in the incremental
change loop increases cumulative system performance, relative to zero delay, but that
increasing this delay to 16 periods makes cumulative performance substantially worse. The
results shown in graphs 3-4 indicate that, contrary to the outcome in the incremental change
loop, no delays in the strategic change loop are beneficial. In particular, recalling Greiner’s
assertion that companies fail during the periods of revolution, it is apparent that delays in
implementing strategic changes lead to very large performance gaps during these periods.
Delay in the strategic change loop increases the risk of failure, and longer delays make
failure a virtual certainty. The reason is that companies that delay strategic changes attempt
to correct their performance problems with incremental changes to their current system
which is fundamentally flawed. Incremental changes to a flawed system do not lead to
success.
An interesting question is why delays in the incremental change cycle initially improve
cumulative performance but then, as the delay increases, lead to decreased cumulative
performance. Graphs 5-7 shed some light on this question. These graphs show the effects
of lengthening delays on perceived performance gap (not cumulated), the number of
strategic change experiments implemented, and the number of incremental changes
implemented. The time span of these graphs is shorter than the previous set in order to
show more detail. These graphs also show the effects of lengthening delays in only one part
of the incremental change loop; between readiness for incremental change and start
incremental changes. The pattern of delays, and essential dynamics, are the same as for the
delay between perceived performance gap and readiness for incremental change. The delay
between readiness for incremental change and start incremental changes is set at zero in
graph 5, eight periods in graph 6, and 16 periods in graph 7. The graphs show that, with
zero delay, incremental changes are implemented earlier. But one of our model
assumptions is that incremental changes cannot improve an existing system for more than
a few years (see Frechette and Spital, 1991), so these incremental changes begin to fail and
drive the perceived performance gap larger in periods 20-35. This larger perceived
performance gap invokes strategic changes, with some simultaneous incremental changes to
the emerging new system, that drive performance up and therefore perceived performance
gap down. As the delay is increased to eight periods, incremental changes are implemented
and also begin to fail, but while perceived performance gap is driven to essentially the same
peak at the same time it is lower during periods 20-35. The performance gap invokes
strategic changes, and incremental changes follow these in order to improve the new
strategic system. Perceived performance is then driven to a higher level (lower perceived
performance gap), as the strategic and following incremental changes work together. Graph
7 shows that as the delays are increased to 16 periods, the perceived performance gap is
initially kept lower but then are driven to a higher peak. The significant difference, though,
is in the behavior during the changes. The long delay makes the incremental changes occur
far behind the strategic changes and, although they initially improve the new system, they
soon can improve it no more and begin to fail. This drives the perceived performance gap
higher again, so the cycle of strategic changes starts again much sooner. Thus the
wavelength of the oscillations is dramatically reduced by this long delay, and cumulative
performance is much worse.
Graph 5. Zero delay before “start incremental changes" Graph 6. 8 period delay before "start incremental changes”
aa ; Lo i
: » ZN
Moe} fT
ae
_ 30.00 X
0.00 35.28 50.80 7818 101.00) ziazpt z Sime
a9 wes any Tes Tore
i
\
to
Graph 7. 16 period delay before "start incremental changes”
Hs
ss aee
Graphs §-7 output variables:
1 Number of strategic changes
la fa
3 Perceived performance gap 4] \ /
2 Number of incremental changes
101.00
Conclusions
Organizations appear to progress through a series of cycles over time. Our simulation
shows such cycles, as would be expected in a negative feedback model with delays. In our
model, changes are invoked by a deviation of actual organizational performance from
desired performance. Incremental changes improve the current strategic system. Strategic
changes create a new strategic system. Periods with high levels of strategic change are what
Greiner has called periods of revolution. Our model outputs illustrate high risk of
organizational failure during these times.
Most discussions of organization design do not address behavior in the context of these
cycles. In this paper, we have attempted to understand the impacts of organization design
on long term performance of a business through these cycles. Forrester’s design
recommendations, combined with shared organizational purpose, seem to affect delays in
organizations. Delays in either information flow or resultant action clearly affect behavior
of the system during cycles.
The simulations suggest to us that most organization designs focus too much attention on
information flow within functions, and not enough on information flow between functions.
The fundamental building block of most organization structures is a functional group, which
fragments attention and objectives and moves information more rapidly within functions but
delays information flow across functions. This fragmented purpose and information flow
may encourage incremental change, which may initially have a functional base, but will
clearly discourage strategic changes which are inherently integrated and depend upon cross-
functional objectives and information. Since delays in our incremental change loop help
long term system performance, provided that they are not too long, but any delays in
strategic changes are far more damaging no matter how long they are, then these structures
may be pushing in exactly the wrong direction with respect to where delays are experienced.
We should encourage cross-functional information flow, and work to reduce those delays,
at the expense of intra-functional information flows.
References
Davis, S. and Davidson, B. 2020 Vision. Simon and Schuster, New York, N.Y. 1991.
Forrester, J.W. "A New Corporate Design", Industrial Management Review, 7, No. 1, (Fall
1965), pp 5-17.
Frechette, H.M., and Spital, F.C. "A Model of Organizational Change" , Proceedings of the
1991 International System Dynamics Conference, Bangkok, Thailand.
Greiner, LE. "Evolution and Revolution as Organizations Grow’, Harvard Business
Review, July-August, 1972.
- 189 ~
CC BY-NC-SA 4.0