A System Dynamics Analysis
of the
New England Ocean Cluster House
Martin Schelasin
Haleigh O’Donnell
John Voyer, Ph.D.
voyer@maine.edu
School of Business
University Of Southern Maine
Abstract
Business incubators have received little attention from system dynamicists. The only system dynamics
study of incubators focused less on how to run an incubator and more on how to use them as part of a
national innovation system. The present paper fills part of this gap with a focused system dynamics
analysis of one freestanding business incubator—the New England Ocean Cluster (NEOC). NEOC’s long-
term goals include evolving New England ocean resource industries into an environment more
hospitable to entrepreneurs, and more conducive of innovation, by relying heavily on collaboration
among incubator clients. A system dynamics analysis of the NEOC revealed significant financial
challenges with its business model. It also showed the fundamental importance to the business model
of intense collaboration among clients.
Introduction: Business Incubators
Business incubators come in many varieties (Barbero, et al. 2012; Barbero, et al. 2014). The first
distinction among types of incubators is between not-for-profit and for-profit incubators (Hansen et al.,
2000). Non-profit incubators foster a social purpose; for-profit incubators try to offer their owners
financial returns. For-profit incubators generally seek returns from fees charged to their entrepreneur
tenants for services provided and from investing in medium- to long-term equity stakes in tenants’
ventures (Becker and Gassmann, 2006). Becker and Gassmann (2006), in their examination of
corporate-based incubators, provide a typology adaptable for freestanding incubators:
e Fast-profit incubator: Commercializes noncore technology with the goal of exiting from its
ventures through spin-off and profit making.
e Leveraging incubator: Increases the utilization of internally developed technology by leveraging
it to the market, hoping to exit by integrating the technology back into the core business to
support the corporation’s future growth.
e In-sourcing incubator: Uses technological knowledge to screen external markets for promising
ideas and high-potential start-ups that it might later “spin in” to expand the corporation’s core
competencies. It exits from the ventures by integrating them into the corporation, through
either an existing or new business unit.
e Market incubator: Tries to develop a market for a complementary non-core technology to
increase demand for its own technology and products. It supports the development of
complementary technologies without having a specific potential acquisition in mind.
Business incubators have received little attention from system dynamicists. The only system
dynamics study of incubators (Tepov, 2013) focused less on how to run an incubator and more on how
to use them as part of a national innovation system. The present paper fills part of this gap with a
focused system dynamics analysis of one freestanding business incubator—the New England Ocean
Cluster (NEOC). In the Becker and Gassmann typology, the NEOC would be a hybrid of the fast-profit
and market incubator types.
The New England Ocean Cluster
The model for the New England Ocean Cluster is the immensely successful Icelandic Ocean
Cluster (IOC). The IOC has gained popularity amongst Nordic ocean resource utilizers and made its mark
in the cod industry by creating a collaborative environment that fostered developments that increased
cod utilization to previously unimaginable levels. The NEOC is an example of what Grimaldi and Grandia
(2005) call an Independent Private Incubator:
IPIs are incubators set up by single individuals or by groups of individuals (companies
too may be among their founding partners), who intend to help rising entrepreneurs to
create and grow their business... They invest their own money in the new companies
and hold an equity stake. (Grimaldi and Grandia, 2005)
The New England Ocean Cluster House (NEOHC), based in Portland, ME, is an office building that hosts a
variety of businesses in the ocean resource field. NEOCH houses a number of offices for firms, with
shared collaborative space, and an incubator offering desk rentals for start-ups within the industry.
Further, NEOC offers the option of participating through non-in-house membership. This cluster aims to
promote creative collaboration that is a result of bringing together NEOC members.
NEOC’s long-term goals include evolving the New England ocean resource industries into an
environment more hospitable to entrepreneurs, and more conducive of innovation. Maine, because of
its flourishing seafood industry and its largely unique access to Lobster as a resource, made it a very
good locale for the North American adaptation of IOC. Patrick Arnold, Owner of SoliDG (a port logistics
and management firm) and head of NEOCH development, posits in his strategic outlines that Maine
harvests $450 million in seafood every year, which is linked to 1.5 million jobs in saltwater fisheries
(Arnold, 2015) Given the overwhelming regional prominence of NEOC’s targeted industries, the NEOCH
would appear bound for success. The present paper will outline the strategic composition of NEOC,
present a model representative of NEOC’s revenue generation, and perform sensitivity simulations to
draw actionable conclusions on the soundness of NEOC’s current operational premise.
as waite Equity stakes in
office spinoffs
one as +
office Zo Se
Be sales of office
spinoff aoa
catsbonton
or Moe ice project a
ay Yer rf Capital gains from
A) cone brakeon spun _ kom otter office spinoffs
4 NEC profit-<<— A —______ NEOC Total
Size and scope of New i eee
England Ocean Cluster ¥
(NEOC) Ss expenses 1 IncuDesk ww
4, Cc : cock COP "pont Capital gains from
—— M nel-ventres IneuDesk spins
ea off 4
Available appropriate IncuDesk
‘commercial space —praustomess
ae. 1 Sales of IncuDesk
InicuDeskifnes: _ spinoff equity
_ Spas ——— yy
Equity stakes in
IncuDosk picts
Tanee a
Figure 1. Causal Loop Diagram for the New England Ocean Cluster.
IncuDesk +
Dynamic Hypothesis
Figure 1’s causal loop diagram captures, at its highest level, the NEOC’s dynamic hypothesis for
achieving financial sustainability. It recruits companies that use its larger office spaces and it recruits
companies that use smaller “IncuDesk” spaces. Both types of companies pay fees that are its primary
short-term source of revenue. By encouraging collaboration, the NEOC promotes spin-offs of both types
of companies, taking an equity stake when they happen, making this a form of equity spin off or equity
carve out (Powell, 2010). Later, the NEOC will sell off its equity, which enters the revenue stream as
well. The major limits on the NEOC’s size are the availability of appropriate commercial space, the size
of the expenses incurred in running the NEOCH, and the degree to which it succeeds at encouraging
collaboration and, thereby, spinoffs.
Preliminary Analysis
In terms of strategy dynamics (Warren, 2008), the NEOC has a distinctly complex multi-faceted
structure. Notably, the organization has three, in some ways four, primary revenue generators. These
categories are in-house office rental, in-house incubator desk rental (IncuDesk), out-of-house NEOC
membership, and the most abstract final revenue stream—capital gains from NEOCH-produced venture
or project investment. In terms of a resource-based view, each revenue stream is associated with a
distinct stock and flow structure (which we did not exhaustively include in the model because of lack of
complete information and over-complication).
Table 1 summarizes this stock and flow structure.
INFLOW
STOCK
OUTFLOW
New Office Companies
Office Customers
Office Companies Lost
Determined by:
Total Potential NEOCH Office
Determines:
Variable Office-Allocated
o .
Cc ies (Not modelled
because of lack of market
research and presumed
expansion of this pool during
the years of operation as the
ocean resource and fisheries
industries expand)
Sensitivity to Collaboration
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the exhibited
amount of collaboration in the
NEOCH)
NEOCH Office Collaboration
Sensitivity to Price
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the current cost for
office rental charged by the
NEOCH)
NEOCH Office Rental Cost
Potential Customer Growth
(Effects of word-of-mouth,
advertising, etc. Not modelled
because of inevitable over
complication of the model.)
Underlying Historical |OC Rate
for Baseline
Operating Revenue from
Offices
Ventures Produced by Offices
Projects Produced by Offices
Collaboration in Offices
(Network Effects: Not modelled
because of inability to
adequately approximate
necessary functions; would
account for the rate of
collaboration’s sensitivity to
more or less offices in-house,
LE. at 0 offices, there can’t be
any collaboration.)
Determined by:
Sensitivity to Collaboration
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the exhibited
amount of collaboration in the
NEOCH)
NEOCH Office Collaboration
Sensitivity to Price
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the current cost for
office rental charged by the
NEOCH)
NEOCH Office Rental Cost
Underlying Historical |OC Rate
for Baseline
New IncuDesk Companies
IncuDesk Customers
IncuDesk Companies Lost
Determined by:
Total Potential NEOCH
IncuDesk Companies (Not
modelled because of lack of
market research and presumed
expansion of this pool during
the years of operation as the
ocean resource and fisheries
industries expand)
Sensitivity to Collaboration
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the exhibited
amount of collaboration in the
NEOCH Incubator)
NEOCH Incubator Collaboration
Sensitivity to Price
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the current cost for
IncuDesk rental charged by the
NEOCH)
NEOCH IncuDesk Rental Cost
Potential Customer Growth
(Effects of word-of-mouth,
advertising, etc. Not modelled
because of inevitable over
complication of the model.)
Underlying Historical |OC Rate
for Baseline
Determines:
Variable IncuDesk-Allocated
Operating Expenses
Operating Revenue from
IncuDesks
Projects Produced by IncuDesks
Collaboration in Incubator
(Network Effects: Not modelled
because of inability to
adequately approximate
necessary functions; would
account for the rate of
collaboration’s sensitivity to
more or less IncuDesks being in
the incubator, i.e. at 0
IncuDesks, there cannot be any
collaboration.)
Determined by:
Sensitivity to Collaboration
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the exhibited
amount of collaboration in the
NEOCH)
NEOCH Incubator Collaboration
Sensitivity to Price
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the current cost for
IncuDesk rental charged by the
NEOCH)
NEOCH IncuDesk Rental Cost
Underlying Historical |OC Rate
for Baseline
New Out-of-House Members
Out-of-House NEOC Members
(OoH Members)
Out-of-House Members Lost
Determined by:
Total ‘ial NEOC bers
Determines:
Determined by:
(Not modelled because of lack
of market research and
presumed expansion of this
pool during the years of
operation as the ocean resource
and fisheries industries expand)
Sensitivity to Collaboration
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the exhibited
amount of collaboration in the
NEOC)
Collaboration in NEOC
Sensitivity to Price
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the current cost for
membership charged by the
NEOC)
NEOC Membership Cost
Potential Customer Growth
(Effects of word-of-mouth,
advertising, etc. Not modelled
because of inevitable over
complication of the model.
Benchmarked on IOC inflow
rates per existing historical
data)
Underlying Historical |OC Rate
for Baseline
Variable OoH: E
to Collaboration
Allocated Operating Expenses
Operating Revenue from OoH
Members
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the exhibited
amount of collaboration in the
NEOC)
NEOCH Incubator Collaboration
Sensitivity to Price
(Approximated in model
because of lack of market
research. This function is in turn
dictated by the current cost for
Membership charged by the
NEOC)
NEOC Membership Cost
Underlying Historical 1OC Rate
for Baseline
New Pending Capital Gains
from or Project Investment
Pending or Project Capital
Gains
Realized Capital Gains
Determined by:
Average Stake in Ventures
Purchased Monthly
Average Value of Ventures
Average Projected Return of
Ventures
Average Stake in Projects
Purchased Monthly
Average Value of Projects
Average Projected Return of
Projects
Determines:
Cumulative Retained Earnings
(In contribution with Net
Operating Income)
Determined by:
User/Company Decision
Table 1. NEOC Stocks and Flows Chart
Model
Translating the above into a model required the dissecting of NEOC into four substructures, one
for each of its revenue streams. The reader will find a shared Sysdea version of the model at
https://a
Figure 2. Office Substructure
.sysdea.com/shared/y6ZN780z5Fu9ICvGkPqzl
For the three
rental/membership revenue
sources (offices, IncuDesks and Out
of House members), the structure
was relatively straightforward. (See
Figures 2 to 4.) The first element
represents the inflow of new
customers, modified by price
sensitivity, collaboration sensitivity,
and capped at the maximum
capacity when applicable with a
first-order control. This inflow
drives the stock, which then acts as
a multiplier base for incremental
variables, such as revenue per
offices or desk or member, projects
or ventures per month per office or desk, and costs per office or desk or member. Meanwhile, much like
the inflow, price sensitivity and collaboration sensitivity drive the outflow. Importantly, there are
additional holding stocks that represent the resource of produced ventures. These stocks drain into the
fourth, non-rental or membership substructure that calculates pending capital gains, and their
respective contribution to revenue inflow.
Ss
Figure 4. Out-of-House Membership Substructure
\
Figure 5. Investment Substructure
Pertaining to the final of
the four major substructures, the
investment or spin off
substructure, the holding stocks
from the office ventures/projects
and incubator projects flow
through equations that modify the
degree to which NEOC holds equity
investments in them. Specifically,
the projected yield of the
investment influences the amount
invested; the model converts the
yield to a dollar amount using an
average of project value. The
converted dollar value of capital
gains then collects in a cumulative
pool of pending capital gains that
can be “realized” via an outflow
dictated by a user or company
decision. (See Figure 5)
All of the revenues from
these substructures run into a
cumulative revenue pool, which
nets through sunk expenditures
and non-allocated recurring
expenses. This allows for the final
calculation of net equity to date.
Simulation Runs
The simulation runs reported here aimed to analyze one primary facet of NEOC: the effect of
collaboration on the revenue of the organizatio
n. This is one element of what Bergek and Norrman
(2008) call “strong business support,” and is the primary objective of what the NEOC wants to provide its
tenants and members (Arnold, 2015). Collaboration is an extremely dynamic and hard to simulate
element of the NEOC’s strategic considerations
collaboration itself, as well as an absence of mai
because of the absence of absolute measures for the
rket research to adequately gauge the target
demographic’s sensitivity to collaboration when making rental or membership decisions.
————
——$—$———
Ce
Figure 6. Simulation 1 Results
considerably smaller than the average possible
Simulation 1: Base Case
To analyze the effect of collaboration on the
revenue stream of the NEOC successfully, we first ran
through a simulation using the best base-projections
possible: the historical numbers reflective of the IOC. To
represent a standard/average amount of collaboration
we used a value of .5, when applied to the underlying
functions for sensitivity to collaboration; it produces the
lOC-derived base numbers. Because of the intended
isolation of Collaboration as the experimental variable,
we left static all other decisions concerning price, at the
rates disclosed by NEOC managers.
Figure 6 shows the results of simulating the first
five years of operation. For a concise outline of relevant
metrics, see Table 2, where we can see that the
simulation assesses that NEOC will be unable to reach a
break-even point even after five full years of operation.
Originally, a two-year break-even had been predicted,
prior to the final leasing of a location for the NEOCH
size evaluated beforehand (final location 16 office and 10
desk capacity, original projection of 30 offices 12 desks). This break-even, however, may not include all
possible credits and or additional revenue streams. NEOC management alluded to a large revenue
Metric
Value at end of Simulation
Net Retained Earnings to Date
(Cumulative Net Income less Sunk Expenditures)
-$118,724.96 USD
Pending Capital Gains
(Standing Unrealized Value from Investments,
not accounted for in Net Retained Earnings to
Date)
$37,529.03 USD
Office Members
Effectively Maximized at ~16
Maximum reached after 13 months
IncuDesk Members
Effectively Maximized at ~10
Maximum reached after 5 months
Out of House Members
33.7
Table 2. Simulation 1 Core Metrics
stream that helped cover incubator costs; this may roll over to cover other expenses and shift the
breakeven forward dramatically.
Other important points of interest from the base case simulation include the amount of Pending
Capital Gains (effectively unrealized revenue not accounted for in net revenue), the number of out of
house members, as well as the number of months it took to reach capacity in both the offices (13
months) and the incubator (five months).
Regarding the pending capital gains, the base case simulation projects that, after five years of
operation, the NEOC will hold $36,529.03 worth of equity in projects and ventures from the house and
incubator, assuming it realizes none during this time. This metric is of critical importance in maintaining
a comprehensive view of NEOC’s all-inclusive value and income.
The number of Out-of-House members represents the only incremental revenue stream for the
NEOC that physical capacity does not affect. As such, its variation because of shifts in collaboration
represent, arguably, the most important factor NEOC’s management can manipulate to expedite their
break-even process. During the base simulation, the number of out-of-house members reached 33.7.
Addressing the rates at which the NEOC reaches capacity maximums, this is important as it is
simply an indirect representation of gained or lost revenue. In the base simulation, the NEOC reached
capacity in the offices after 13 months, and in the incubator, in only 5 months. If we assume each month
represents lost incomes in the offices and incubator of approximately 1.5k * (empty offices) and 250 *
(empty desks) respectively, the considerable impact on revenue becomes dramatically apparent.
10
Figure 7. Simulation 2 Results: Increased
Collaboration
(Base case represented by non-bold lines)
Simulation 2: High Collaboration
Progressing from the base case, we ran the
second simulation with a hypothetical collaboration
rate of .8, approximately 160% the collaboration of
the base case. (See Figure 7) The aim of this
simulation was to deduce whether it would be
effective, and financially viable, for the NEOC to
encourage even higher volumes of collaboration.
After simulating the first five years of
operation (60 months), we show the charts for the
produced outputs in Figure 7. For a concise outline of
relevant metrics, see Table 3, which shows that, even
with a considerably higher rate of collaboration, it is
readily apparent that NEOC fails to break-even within
a 5-year window (absent any extenuating unknown
factors as previously mentioned). However, the
amount of underlying debt from sunken expenditures
is considerably lower, and is nearly counterbalanced
in full by a doubled amount of pending capital gains in
contrast to the base case; it increased by $37,738.47.
(See Figure 8)
In direct reflection of the increased
collaboration rate, the points at which office and
incubator spaces reached their caps moved forward.
While the increase was somewhat negligible in the incubator, in the case of offices the move forward
represents multiple thousands of dollars in gained revenue.
Lastly, the collaboration increases effect on the number of out-of-house members was
stupendous. It increased the amount at simulation end to nearly 150% (148.07%). Because of the
miniscule fees required for membership this ultimately led to small shifts in revenue and a largely
unnoticeable effect on breakeven. The implications of this however, are immense should NEOC use this
information as a propeller to assess potential changes in OOH membership fees.
Metric
Value at end of Simulation
Net Retained Earnings to Date
(Cumulative Net Income less Sunk Expenditures)
-$83,609.78 USD
Pending Capital Gains
(Standing Unrealized Value from Investments,
not accounted for in Net Retained Earnings to
Date)
$75,267.50 USD
Office Members
Effectively Maximized at ~16
Maximum reached after 4 months
IncuDesk Members
Effectively Maximized at ~10
Maximum reached after 3 months
Out of House Members
49.9
Table 3. Simulation 2 Core Metrics
11
Figure 8. Simulation Comparison: Base versus Figure 9. Simulation 3 Results: Reduced
Increased Collaboration Collaboration
(Base case represented by non-bold lines) (Base case represented by non-bold lines)
Simulation 3: Low Collaboration
To affirm the conjectures supported by the high collaboration simulation, a proof-of-concept
test lowering collaboration is of equal or greater value because of its inherent creation or disproval of
the observed pattern. This investigates the primary contrasting position of that in Simulation 2; NEOC
loses such negligible revenue in a low collaboration environment that initiatives to increase
collaboration are not only not profitable, but a noteworthy misuse of resources. To simulate this
scenario, we used a collaboration rate of .3. We show the output for simulating the first five years of
operation in Figure 9. For a concise outline of relevant metrics, see Table 4.
In dramatic contrast to the prior simulations, a decrease in collaboration of even 40% results in
catastrophic breakdown of the NEOC’s revenue structure unless remedied aggressively. After 20 months
of clinging to profitability, the lack-of-collaboration’s effect on member/office/IncuDesk outflow
overwhelms the point of inflection for profitability, and the standing debt of the NEOC increases steadily
from there on. This supports the operating theories introduced by Simulation 2. NEOC would not only
benefit immensely from spurring additional collaboration, but it faces inevitable failure in the event it
does not maintain collaboration in the immediate range of the initial |OC benchmark rates.
12
Metric
Value at end of Si
Net Retained Earnings to Date
(Cumulative Net Income less Sunk Expenditures)
-$325,719.67 USD
Pending Capital Gains
(Standing Unrealized Value from Investments,
not accounted for in Net Retained Earnings to
Date)
$17,576.87 USD
Office Members
Fails to Reach Maximum
Ending Value of 11.6
IncuDesk Members
Effectively Maximized at ~10
Maximum reached after 9 months
Out of House Members
24.1
Table 4. Simulation 3 Core Metrics
Known Oversights & Model Caveats
Table 5 summarizes unaccounted-for variables pertinent to the model.
Consid.
‘ion & Elaboration
Adoption Rates & Finite Potential Customer Pool
This model does not currently account for a finite
pool of potential customers for any of the
operating areas. This may or may not be a valid
assumption going forward depending on growth
rates within concerned ocean resource
industries. Additionally, it does not model
“adoption rates” such as the effect of word-of-
mouth on the inflow and outflow rates within the
NEOC.
Factors Affecting and/or Calculation Process for
Collaboration
Collaboration is represented as a user/company
defined quantified variable. This is may be
considered a brash assumption and may be
better suited to having a sub-model which can
better approximate a value. (The substructure
would likely be based on project or venture
output, but would then be subject to risk of a
circular reference interfering with proper
calculation)
Pre-Tax Revenue Output
The model does not currently adjust for taxes and
as such is an approximation.
Percentage Investment Sensitivity to Project or
Venture Size
This model currently does not alter the amount
invested in projects/ventures based on their size.
A variable for available liquid capital would likely
be needed to disallow investing more money
than the NEOC had on hand, and limit its
investment percentage in extremely high size
projects. (I.E. if through some means a $100m+
project was produced, it is highly unlikely NEOC
13
would approach it with the same investment
mentality as a $1m project)
Compounding CG’s Does not account for the compounding of capital
gains after their purchase over multiple years of
holding. Presumably, modelers could fix this with
multiple substructures for each project in which
there is an investment, and with an accumulating
stock that tracked the compounding amounts.
Table 5. Model Oversights & Caveats
Summary & Closing
The simulations presented in the present paper show both an intricate elaboration on the
NEOC’s revenue structure and a functional hypothesis that collaboration stands as one of, if not the
most, critical element to the NEOC’s longevity and profitability.
The base case in Simulation 1 presented a scenario in which the NEOC reached Icelandic Ocean
Cluster levels of collaboration, venture production, project production, and inflow and outflow in offices
and the incubator. This case offered a solid case for NEOC being a profitable venture in the long-term,
even without strategic shifts in the current operational layout.
In Simulation 2, evidence arose of a critical link between NEOC’s approach to break-even and
collaboration in the incubator and office environments. Not only did the revenue in the NEOCH increase,
but also the number of out-of-house members of the NEOC increased dramatically. This led to the
conclusion that increases in collaboration, coupled with an increase in the cost for out-of-house
membership, would better NEOC’s ability to achieve break-even at an attention-demanding magnitude.
Simulation 3 confirmed the pattern suggested in Simulation 2 and precipitated a pressing
change in implication. The hypothesis shifted from support for the value of increase in collaboration, to
include the strict inadvisability of allowing collaboration to drop in any considerable capacity.
Lastly, we can recommend a number of operational methods to the NEOC. These
recommendations are likely generalizable to many Independent Private Incubators (Grimaldi and
Grandia, 2005) that seek to use collaboration as their primary contribution to the enterprises they host:
- Promote collaboration in any, and all, ways possible within reasonable cost bounds.
- Develop a proprietary formulaic approach to calculating the rate of collaboration in the
house, and the cluster as a whole.
- Actively track and maintain running reports on the rate collaboration in the house (as
calculated using the new formula); performing routine analytics to identify warning signs in
this data is of critical importance.
- Create a scalar model to dictate increases in membership price as collaboration increases or
decreases (this is theoretically accomplishable by modification of the model supplied in the
present paper).
The NEOC will likely find success in its current format. However, should it approach its first years
of operation with the above-mentioned tactics enacted, it will reach break-even at a significantly faster
rate, and will avoid undue risk to its continued operations.
14
REFERENCES
Arnold, P. 2015 Personal Communication, September 29, 2015.
Barbero, J.L., Casillas, J.C., Ramos, A. and Guitar, S. 2012 “Revisiting incubation performance: How
incubator typology affects results.” Technological Forecasting & Social Change, 79: 888-902.
Barbero, J.L., Casillas, JC, Wright, M and Garcia, AR. 2014 “Do different types of incubators produce
different types of innovations?” Journal of Technology Transfer, 39:151-168
Becker, B. and Gassmann, O. 2006 “Gaining leverage effects from knowledge modes within corporate
incubators.” R&D Management, 36, 1: 1-16.
Bergek, A. and Norrman, C. 2008 “Incubator best practice: A framework.” Technovation, 28 (1-2): 20-28.
Grimaldi, R. and Grandia, A. 2005 “Business incubators and new venture creation: an assessment of
incubating models.” Technovation, 25: 111-121.
Hansen, M.T., Chesbrough, H.W., Nohria, N. and Sull, D.N. 2000 “Networked incubators: hot houses of
the new economy.” Harvard Business Review, 78, 74-84.
Powell, B.C. 2010 “Equity carve-outs as a technology commercialization strategy: An Exploratory case
study of Thermo Electron’s strategy.” Technovation, 30: 37-47
Tepov, R. 2013 “The Impact of National Innovation System on Entrepreneurial Venture Creation
Process.” Master’s Thesis, Lappeenranta University of Technology.
Warren, K. 2008 Strategic management dynamics. Hoboken, NJ: Wiley & Sons.
15
attempts. Since pursuing this option called for understanding the nature of the two capacity
portfolios and managing investments in them, it was seen as a positive learning outcome. The
gaming strategies were however based on the knowledge that the game would end after a finite
play period and strived to manage the timing of the investment portfolios rather than devising a
state-based strategy. In a real situation, a state-based strategy would be the only sustainable
option. More discussion on the merits and demerits of the time and state based strategies was
needed, but could not be pursued in this first run of the game in view of the time constraints.
An important outcome was that the game players wanted to pursue the links between the various
decision sectors of the game, which was consistent with the intended objectives of the game.
They in fact helped to draw these links on a flip chart available in the room that allowed the
group to discover the map of Figure 3 and give a meaning to their suspicion that knowing these
links is important to devising a successful winning strategy.
There was time enough to point out that while both plane seats and service capacity portfolios
affected ridership, the investment decisions in them were made on very different bases. The
plane-seats were a visible and tangible portfolio and investment in it could be linked to the
revenues, which created a clear basis for determining the size of the fleet. The service capacity
on the other hand resided in many elements, including passenger handling, scheduling, baggage
handling, counter and on-board services, training, corporate culture, problem solving capacity,
etc. Its need could be assessed only when there was a perception of its shortage motivated by
customer complaints and management observation. Thus, the fleet expansion decisions were
driven by a positive feedback whose gain could be controlled, the capacity expansion decisions
were driven by a negative feedback that arose out of a perception of discrepancy that took a
while to come to fore. The first portfolio could be pro-actively managed, the second could not
be.
The state-dependent resolution of this problem requires re-examining the information links and
decision rules and therefore modifying the model, which can be attempted either by a trained
modeler or through extended discourse which could not be pursued in the available time. The
time dependent resolution of the problem required monitoring the aggressiveness of expansion of
Page 17 of 21
each capacity portfolio over time carefully by continuously changing slider settings provided in
the control panel. This option was pursued by most groups - with success.
Last but not least, the debriefing discussion led to recognizing a correspondence between the
portfolio management problem in People Express and the VA homeless program where
homelessness prevails in spite of increasing shelter provision. Some could also see that the
imbalance between tangible and intangible service portfolios also prevailed in other VA
operations. The take away from the game was clearly the need to manage these portfolios and the
references to the game were quite prolific in the subsequent day deliberations of the meeting
participants on their operational problems.
Our debriefing followed the spirit of structural debriefing outlined in Pavlov and Saeed (2015),
although due to time limitations no attempt was made to group-model the game-play, which we
highly recommend in a more drawn out exercise.
Conclusion
This strategic thinking exercise enabled VA senior executives and managers to confront many of
the same challenges they encounter in the course of their demanding roles in the initiative to end
Veteran homelessness. However, they were able to escape the confines of their existing
individual mental models during the course of this strategic thinking exercise because the
scenario was set in a different sector. The exercise enabled the audience to develop new shared
mental models regarding latent capacity portfolios. Shared Mental Models are the building
blocks of Organizational Learning, which increases the organization’s capacity to take effective
action (Kim, 1998). These mental models are now enabling executives and managers across the
VA Homeless Programs to make better resource allocation and operations management decisions
by considering service capacity alongside bed capacity. Since these mental models are shared,
there has been broad support for decisions to limit the rate of bed capacity growth until service
capacity issues are adequately addressed. The understanding that service capacity is mostly
hidden and hard to measure has also informed Operations Management efforts. This has been
Page 18 of 21
especially important for those VA Programs that have seen rapid expansion in shelter capacity
over the past few years.
The People Express scenario and the structure of the game were highly relevant to the
participants because they were able to relate key elements to their own organizational roles. VA
Homeless Programs have experienced recent growth in physical capacity. In the weeks leading
up to the conference, VA was undergoing a leadership change and the overall VA strategy was
evolving. Customer service was becoming a key focus area for the new VA leadership. Lastly,
similar to the structure of the game, participants faced an emphasis on achieving the strategic
target of ending Veteran homelessness in a fixed time frame.
While these results exceeded the expectations for the exercise, participants also walked away
with additional key insights that are highly relevant to VA Homeless Programs Operations.
Participants underscored the need to relate to long-term targets in terms of short-term outcomes
during the course of efforts to end Veteran homelessness. Another lasting insight was the
importance of anticipating system level outcomes in efforts to sustain progress made in Veteran
homelessness. Lastly, participants embraced systems thinking and resonated with the need to
fully explore and understand the underlying drivers, delays and feedback loops leading to system
level outcomes. Many months after the exercise, senior managers continue to draw references to
the people express scenario and insights gleaned from the exercise.
Such applications of serious gaming to promote strategic thinking require the support of
executives who recognize the pitfalls of using conferences as social events that discuss little
more than pending tactical issues. Leveraging serious gaming to promote strategic thinking not
only promotes connections across the organization but can also serve as a forum to tackle thorny
strategic challenges. Such exercises are another tool for organizations like VA that are striving
towards agile operations and continuous learning (VA, 2014). However, greater adoption of such
tools will require attention to both their cost and realized value. This effort realized significant
costs savings by using a well understood generic structure. A greater awareness of off-the-shelf
models and in-depth understanding of the applicability of generic structures across domains can
lower the cost of developing such exercises.
Page 19 of 21
The three-day conference devoted a whole afternoon to the strategic thinking exercise. While this
was a substantial investment on part of executives and managers, time constraints limited the
value that could be tapped. For example, upon conclusion of the de-briefing session, participants
wanted to play another round and apply newly gleaned insights into the nature of complex
systems. This second round would have helped in consolidating systems thinking and reinforcing
new mental models that were taking shape. In addition, scheduling and funding constraints
limited attendance at this conference, as they do at most conferences. Participants believed that
the strategic thinking exercise was relevant to many of their peers and could have benefited five
times as many participants. Many participants requested a copy of the game and intended to
further explore the exercise when they returned to their offices.
While such independent use of the game could realize additional value, it also presents new
challenges and limitations. Independent use of the game does not have the benefit of a game
facilitator to guide the participants in the process of converting their experiences into insights.
Secondly, independent usage is void of group dynamics and cannot fully leverage game
mechanics. Lastly, use of the game outside the confines of a structured exercise presents the risk
of unintended applications that stretch the model beyond its intended usage. Despite the
successes of this strategic thinking exercise and the willingness of the organizers to devote
additional time in future events, significant potential remains both to scaling both the depth of
engagement and breadth of participation in the strategic thinking exercise. The advent and
maturation of social gaming technologies presents significant opportunities for achieving this
scale.
References
Department of Veteran Affairs (VA). 2014. Blueprint for Excellence
http://www.va.gov/HEALTH/docs/VA_ Blueprint for Excellence.pdf
"
Eisenmann, Thomas R., and Lauren Barley. "People Express Airlines." Harvard Business School
Case 812-134, April 2012.
Forrester JW. 1961. Industrial Dynamics. MIT Press. Cambridge, MA.
Forrester JW. 1969. Urban Dynamics. MIT Press. Cambridge, MA.
Page 20 of 21
Forrester JW. 1971. World Dynamics. Wright-Allen Press, Cambridge, MA.
HUD. 2014. The 2014 Point-in-Time Count. U.S. Department of Housing and Urban
Development (HUD)
Kim, Daniel H. The link between individual and organizational learning. In The strategic
management of intellectual capital (1998): 41-62.
Mullen, Joan and Leginski, Walter. 2010. Building the Capacity of the Homeless Service
Workforce The Open Health Services and Policy Journal, 3: 101-110
Pavlov, O., Saeed, K., Robinson, L. 2015. Improving Instructional Simulation with Structural
Debriefing. Simulation and Gaming. doi: 10.1177/1046878 114567980
Saeed, K. 2009. Stray dogs, street gangs and terrorists: manifestations of a latent capacity
support system. Proceedings of 2009 International System Dynamics conference.
Albuquerque, NM: System Dynamics Society
Sterman, J.D. 1988. People Express Flight Simulator.
http://www.economia.uniparthenope.it/modifica_docente/ferretti/CASO GEOX PEOPLEE
XPRESS.PDF
Sterman, J.D. 2000. Business Dynamics. lwin Mcgraw-Hill
Page 21 of 21
