The application of System Dynamics to the management of a small firm?
A Case of Study of the wine industry: C antine Settesoli
Enzo Bivona
Candidate Master of Phil. in System Dynamics
University of Bergen, Norway
e-mail: enzo.bivona@ ifi.uib.no
Abstract
This paper analyzes a winery company using System Dynamics to investigate policies
for sustainable development. The wine production sector and the world market show
complex characteristics that limit real understanding of the behaviors generated.
System Dynamics can be a powerful tool to improve understanding and find leverage
points of complex management systems by supporting managers testing "what-if"
scenarios and exploring what might have happened - or what could happen - under a
variety of different past and future assumptions and alternative decisions. The
company strategy for penetrating world wine market meets resistance factors - outside
- inits competitors, but as well - inside - in its own structure.
This System Dynamics model improves understanding of the real factors limiting
company performance and helps decision makers to overcome company constraints.
Introduction
Cantine Settesoli! is an Italian co-operative that produces red and white wines of
different qualities. Its challenge is to increase market share in Europe and to penetrate
USA and China markets. It sells about 50% of bottled wine in Sicily, 5% in Northem
Italy and exports 45% to Europe. The model (see fig.1) investigates the relationships
between the wine market and Cantine Settesoli on a side, and between the company
and its partners on an other.
The market segment to which Cantine Settesoli belongs (holding a relatively low
market share) is characterized by a high competition among producers and a
fragmentation of offer. The firm sells its product both to distributors (who bottle and
' Due to confidentiality reasons, figures reported in this paper have been disguised.
re-sells it with their own trade mark) and to other producer firms who may need more
wine to satisfy their market requests.
Wine Market
Bottled Sector | | Cantine Settesoli
ad Partners
Non-Bottled Sector a canting
Other.... |
Fig. 1 - Cantine Settesoli model overview
Bottled wine is sold at a higher price than the non-bottled one.
The firm produces and bottles wine made from grapes supplied by several small
farmers, who are its partners.
The average life time of its grape plants is fifteen years: this is a major constraint to be
taken into consideration when business plans are drawn up.
Currently, land devoted to white wine production is significantly larger than the one
for red wine production and total supply does not match market demand at all. The
size of the co-operative and the “local environment” where it operates allows business
decision makers to find a product mix that meets market demand. So, in order to
match such demand, the firm is forced to invest more financial resources to buy a
relatively expensive red wine or grapes from external suppliers. Such an extra-cost
that the company has to sustain is perceived by the management as a significant
constraint for future growth.
Problem Definition
A System Dynamics model has been sketched in order to analyze cause and effect
relationships between relevant variables related to the dynamics of supply and
demand, with a view to support land reallocation policies aimed at satisfying market
demand. A main hypothesis on which the model has been built is that a higher grapes
yield would accelerate land re-allocation policy by partners.
Should the company performance improve, if current production will be shifted from a
70% white vs. 30% red wine to a more balanced product-mix? Which consequences
could produce the above decisions on the partners’ yield?
Counterintuitive behavior of the system underlying such processes and perils related
to a policy aimed at postponing such production shift, were the major reasons
requiring the use a System Dynamics model.
Cantine Settesoli Model
Figure 2 illustrates system boundaries and the model's sectors.
Not Included Other Segments
te |
| Exogenous Ratio Quality/Price |
| Industrial Sector |
Competitors |
| Endogenous |
I
| Land Sector
| Bottled Production Sect
| Market roduction Sector] |
| Demand Salesmen Sector |
| Financial Sector |!
Fig. 2 - Cantine Settesoli model boundaries
The model consists of four main sectors:
1. The land sector describes how total land available is devoted to the production of
red and white wine. It shows the grape planting cycle (graph 1, page 4): every
fifteen years grape plants die and need to be replanted. It takes three years to
“regenerate” the land and, hence, to start with a new plant.
2. The production sector describes grapes transformation, aging time and bottling
processes. The hectoliters of wine sent to the Bottling sector depend on
Distributors Demand and Desired Inventory. If the quantity supplied from the
company’s partners is not enough to satisfy demand, as it happens for the red
wine, top management decides to purchase wine or grapes outside.
3. The salesmen sector describes the effectiveness of the salesmen in acquiring new
distributors (clients) and how they maintain these relationships. In relation to
company annual sales target, decision makers hire human resources to visit a
desired number of Distributors. The stock and flow structure to this sector shows
how the distributor loyalty changes wine sales related each age class.
4. The financial sector shows the effects of different strategies on company
revenues, and partners’ yield. It displays industrial costs for processing and wine
bottling, price-quality ratio, revenue and yield for the different varieties of grapes
supplied.
The Land Sector
As previously referred, grape plants have an average life of 15 years. During the first
three years production is relatively lower than normal; at the fourth year it reaches the
maximum level, which remains steady until the 13" year. After that time, the harvest
yield starts to decrease and then (two years later) plants are removed. It is necessary to
wait for three more years to start with a new plantation.
100
Quintal/
Hectares 0
0 T —
o 2 4 6 8 10 12 14 16
year —®—Grapes Plant
Productivity
Graph 1 - Grapes Plant Life Cycle
Figures 3 illustrates main feedback loops related to the white wine production and the
land sector. The reinforcing loop R1 shows how each year the removal of those plants
which are fifteen years old gives raise (after a regeneration time and according a given
yield) to a white land replanting rate. Such replanting rate affects land assigned to
white wine production, which is released after fifteen years.
The balancing loop B1 describes how each year an increasing in land released reduces
the total land available to produce white wine.
A same structure has been used to describe the dynamics of the land devoted to red
wine production.
Z
Land Devoted to White Wine
White Wine Production
tc
White Grapes
Y ‘
Ml Land Land Release
Replanting Rate Rate
. ee
4 Cycle
Land Regeneration
Time
Fig. 3 - White Land Sector feedback relationships
Land Sector Re-A location Policy
The model aims to assess the robustness of a policy leading to a balanced product mix
(i.e.: 50% red and 50% white wine).
Two scenarios have been particularly investigated:
1, Partners resist to follow the Land Sector Re-Allocation Policy.
2. Partners are inclined to re-allocate their land because of the decision of
Settesoli’s policy makers to increase the “Red Grapes Yield” (i.e. to allow them
an increase in the price paid for red grapes supplied).
In the following pages the two above scenarios will be analyzed.
1. Partners resist to follow the land sector re-allocation policy.
Fig. 4 shows main feedback loops related to the first scenario.
+
Land Devoted to ~*~ White Wine
White Wine Production
t+ Prod) ac
White aa Plant Life
wi le
are Nt Land White vant 7,
Land me Rate wa Rate
Regeneration
Comy pany Desired — Roe ite
Red cael fra
“N+ 5 incremen of Lana Red Land
Devoted to Release Rate
Partners -—~¥ Wine aes 2
Resistance Factor 62
t+
Red Grapes Red Land
Yield Planting Rate ena Wine. to
Production
va *
Land ah
Regeneration Red Wine fa
Time
Production
Fig.4 - Land Sector feedback relationships
on
Feedback loop R2 shows how the company desired red grapes area and partners
resistance factor are two parameters which strongly affect the increase in red wine
production. In fact, only if the values of such parameters increase, the land devoted to
red wine production will change and, hence, red wine production will increase too.
Such a scenario, that is related to an unchanged red wine yield, shows the weak and
slow increases of the land devoted to red wine production. Graph 4 shows how land
devoted to produce red and white wine does not reach the company target.
Land_Devoted_Red_Wine_
~T™ Production
Land_Devoted_White_Wine
~27 Production
3+ Company_Target
Hectares
0 5 10 15 20 25
Time
Graph 4 - Land assigned to Red and White Production.
From a combined analysis of the above structure and behavior, it is possible to
observe how the dominant loop is R1. In fact, maintaining the status quo implies that
a higher percentage of land is always devoted to white grapes (i.e. loop R1 is fed
according to the same amplitude) and a smaller size of production is devoted to red
grapes. In other words, loop R1 prevails on loop R2.
2. Partners are inclined to re-allocate their land because of the decision of Settesoli’s
policy makers to increase the “Red Grapes Yield”.
The second scenario shows how the behavior of the two curves related the two kinds
of production is the same as in the previous simulation, as only in the 7.5" simulation
year”, the company policy makers decide to increase red grapes yield. Consequently,
the company’s partners perceive such an yield more as profitable and, hence, start to
shift their production mix towards white grapes plantation. Graph 5 shows how a
higher red grapes yield facilitates a faster shift in production, which allows the firm to
meet the 50% target in about 23 years.
Land_Devoted_Red_Wine_
“1 Production
Land_Devoted_White_Wine
~2~_ Production
Company_Target
Hectares
0 5 10 15 20 25
Time
Graph 5 - Land assigned to Red and White Production.
+
Land Devoted to. “A White Wine
White Wine Production
Production
White Grapes Plant Life
an wi
<S White Land
White Land 2
Land a Planting Rate Released Rate
Regeneration
Time 5
@~ Total mitt
Released Rate ~
Company Desired
Red Grapes Area
+ Incremen of an “\ Land
Devoted to Red Release Rate
Partners __>-¥ Wine Production 2 2
Resistance Factor (@
‘i t+
Red Land
Red Grapes Planting mt Land Devoted t6
Yield Red Wine
Land
Regeneration Red Wine
Time Production
Fig. 5 - Land Sector feedback loops including Management Lever
The rise in red grapes yield, decreasing the partners resistance factor, facilitates the
increment of the land devoted to red wine production. The higher red grapes yield
shifts the feedback loop dominance from R1 to R2.
? Another simulation, that has not been included in this paper (due to a page limit), shows how if the
company starts to foster such a shift in plantation since the first simulation year, an even better
performance is achieved.
Company Desired Red Grapes Area
a
AP 6 ren et
Red Grapes Yield ,
aS Tol Land Avelable re
a Partners_Resistance_Factor Tk
A penne of_Red we) 2
an
Desired Increase Red Land
Land_Devoted to Red Wine Productio ——
oma —e™r
fed Land Ping Re
—=<x—
(Whiteland Release #
/ | Yearstit Pens)
/ sage) / \ \
/ { >|
{Ystit ps) |
( rons)
i?) a \ x ee |
sere Ro ed es \\ reseeate tie \, /
TeaLlnanie erin CS \ esis! /
CO) \ } \
i x i of i,
NE A Aa
Planting Rate Land Released
Fig. 6 - Stock and Flow Diagram Land Sector
Figure 6 shows the land sector stock and flow diagram. It particularly portrays how
the two stocks of the land devoted to red and white wine production are affected by:
¢ the planting rate;
¢ the release rate,
¢ the aging rate of plants,
* partners resistant factor (affecting the increment of red land),
¢ red grape yield (affecting partners resistance factor),
* total land available.
Main financial outcomes related to the two scenarios
Financial consequences of these above two scenarios are shown in graph 6, which
displays the different pattern of grapes yield. The higher value of curve 2 demonstrates
how the second scenario leads to a higher total grapes yield. In fact, an increase of red
wine production allows the company to reduce red wine purchase costs from outside
suppliers and, as a consequence, partners may earn a higher yield.
Total Grapes Remuneration
2
mr
1 a aii
$ a ee
101 ae
12
4
Graph 6 - Total Grapes Y ield
Legend:
Line 1 shows partners yield in the case of resistance to the re-allocation policy.
Line 2 shows partners yield in the case in which policy maker increase the red grapes yield to facilitate
the re-allocation policy.
Conclusions
From the above considerations it emerges how System Dynamics may give a major
support to a better understanding of interconnected relationships between the different
company systems and policy making.
“The behavior of complex systems is very often surprising, even when we are fully
aware of the basic interdependencies within ...... systems” (Forrester 1971).
A holistic view allows top management to plan and test “correct” policies (including
time and necessary resources to invest) for the business system, so facilitating the
achievement of company target.
This paper brings together insight and contributions of many people. But a special
note of thanks goes to prof. Carmine Bianchi and prof. Paal Davidsen.
References
Davidsen P.I. (1991) “The Structure-Behavior Graph: Understanding the
Relationship Between and Behavior in Complex, Dynamics Systems,” Department of
Information Science, University of Bergen, Bergen, Norway, System Dynamics
Group, MIT, Cambridge. USA.
Forrester, J.W.1968. Market growth as influenced by capital investment; The
Industrial Management Review, Vol.9, No.2, pp.83-105, Winter 1968b.
Forrester, J.W. 1961. Industrial Dynamics. Productivity Press. Portland, Oregon.
USA.
George P. Richardson, Alexander 1. Pugh III. 1981. Introduction to System Dynamics
Modeling with Dynamo. Productivity Press. Portland, Oregon. USA.
Amulf Grubler. 1997. Time for a Change On the Patterns of Diffusion of Innovation.
Engineering Management Review. IEEE. Summer
10
