Resolving Performance Measure C onflicts in a Supply Chain
Using Systems Thinking Methodology
Kambiz Maani
The University of Queensland, Australia
k.maani@uq.edu.au
Annie Fan
The University of Auckland Business School
ABSTRACT
Performance measurement and management have received a great deal of attention in
the literature in recent years. However, to date, there is scant attention to dynamics
and trade-offs amongst performance indicators in theory and in practice (Santos,
Belton et al. 2002). Thus, performance management systems (PMS) have remained
static, fragmented, and backward looking (Boume et al. 2000) leading to adverse
outcomes, often unknown to managers and organisations. A systems view of
performance, on the other hand, calls for a holistic approach to performance
measurement integrating multiple dimensions, functions and time horizons across the
enterprise. A systemic performance measurement would take into account the
interdependencies of functions and their dynamic influence on the performance of the
organisation as a whole. This paper addresses this challenge using the four level
thinking (Senge, 1991) and causal loop models to highlight the inter-relationships
between the KPIs and their trade-offs within and across different functions. The study
reports on an action research within a multinational company where through real case
scenarios we demonstrate how KPIs influence, contribute or impede one another in a
manufacturing/supply chain setting. The paper reveals how the use of systems
thinking concepts and causal loop models by novice users facilitated an open
environment for cross-functional communication and collaborations, leading to team
and organisational learning and enhanced performance.
Keywords: performance measurement, team learning, mental models, systems
thinking, cross-functional management
INTRODUCTION
“Not everything that can be counted counts and not everything that counts can be
counted.” Albert Einstein
Numerous frameworks have been proposed to help organisations define indicators
that reflect their objectives and assess their performance. Examples include
Performance Pyramid (Lynch and Cross 1991), the Balanced Scorecard(Kaplan and
Norton 1992) and Performance Prism (Neely, Admans et al. 2002). Recent literature
indicates a shift from treating financial performance as the foundation for
performance measurement to treating them as one among a broader set of indicators.
The new performance measurement frameworks aim to connect performance
indicators to business strategy and are designed to be multi-dimensional, explicitly
balancing both financial and non-financial measures, both leading and lagging
indicators to overcome the limitations of the traditional financial measurement
systems. However, a key criticism of current performance measurement frameworks
is their static nature (Todd 2000).
KPIs are widely used by organisations to track actual performance against targets to
assist decision making. Although the use of KPI is prevalent, there remains an
underlying complex problem of correctly identifying and addressing trade-offs
between a set of KPIs. Maani and Li (2004, 2005) suggest that too many KPIs could
lead to over-reactions and over-intervention with adverse unintended consequences
for organisations. This is because KPIs are often viewed as ‘linear’ - without paying
due attention to interactions amongst them. In addition, far too many organisations
still define their performance measures without understanding the dynamic
interdependencies and trade-offs between the individual or groups of indicators
(Santos, Belton et al. 2002).
The existing literature on performance measurement and evaluation appears to have
overlooked the critical dimension of trade-offs between the performance indicators
and strategic objectives. Trade-offs are inherent in complex systems - in particular in
a business, social and policy environments. Understanding the dynamic interactions
between KPIs allows decision makers to prioritize conflicting interests and objective
and to achieve greater enterprise wide result.
LITERATURE REVIEW
The prevailing performance management systems tend to be driven by short term
goals and local optimisation (Neely, P et al. 1999; Y oungblood 2003), discouraging
continuous improvement and learning (Lynch and Cross, 1991) and lack external
focus (Kaplan and Norton, 1992).
Catellano, Young et al. (2004) identified seven fatal flaws of performance
measurement outlines below:
Ignoring the Performance Contributions of Interactive System Elements
Misunderstanding V ariation
Confusing Signals with Noise
Misunderstanding Psychology
Confusing the V oice of the Customer with the Voice of the Process
Failure to Support a Process View
Misunderstanding the Real Role of Measurements
SAO PE Na
Most organisations and managers, by extension, treat different elements of
performance as independent and in isolation. This stems from a lack of systemic and
integrated view of their organisational units leading to silo mentality and internal
competition. "Unless performance management has an enterprise scope, an
organisation cannot synchronise measurement across departments and gain true
visibility of business performance” (Bourne, Mills et al. 2000; Castellano, Y oung et
al. 2004).
Furthermore, the real role of performance measurement is often misunderstood. Often
managers are so consumed with lengthy data gathering and mindless micro
management that they lose sight of broader organisational objectives and strategy.
This could lead to proliferation and over-complexity in performance measurement
systems (Johnston, Brignall et al., 2002).
Performance Trade- offs
The concept of trade-offs is not new. However, there is no consensus amongst
researchers as whether or not they are avoidable. Skinner (1969) first proposed the
trade-off theory and defined that trade offs are unavoidable in the competitive
business environment. Organisations are constantly competing along multiple
objectives therefore it is argued that higher performance in one objective can only be
achieved by compromising the performance of another. Santos, Belton et al. (2002)
state that trade-offs between performance indicators are inherent in the business
environment. If there are multiple objectives for an organisation, then by definition
they must be conflicting, otherwise there would only be one objective (Y oungblood
2003). Slack (1991) believes that trade-offs exist only in the short run; they can be
eliminated in the long run. However, Silveira and Slack (2001) suggest trade-offs do
exist and they can only be lessened but not eliminated.
Collins (2001) on the other hand suggests that organisations should abandon trade-
offs and replace them with an approach to complement competitive objectives. He
maintains the ultimate aim of organisations should be focusing on satisfying customer
needs by achieving all competitive objectives.
Ferdows and de Meyer (1990) take a middle ground suggesting that trade-offs do
exist but the trade-off theory does not apply in all cases; instead it is a function of
progressive development of each performance dimension upon the others. They state
that in the short term, it is possible to trade off capabilities against one another, but in
order to construct long-term capability, management must develop the four objectives
of quality, dependability, speed and cost in a ‘sand cone’ fashion where the lower
layers must be extended in order to support any increase in any higher layer.
RESEARCH MODEL & METHODOLOGY
The research approach used here is the Four Level Thinking Model (Senge, 1991,
Maani and Cavana 2007). This model consists of four distinct but related levels:
events, pattems of behaviour, systemic structures and metal models (Figure 1). It is
argued that most management and policy actions unfold in this manner, where events
represent the shallowest yet most visible level of reality and mental models reflect
deepest and most profound assumptions, norms and motivations (i.e., individual as
well as organisational culture).
The research methodology employed was action research within a multi-national food
company (referred to here as FoodCom) where one of the authors is employed as the
supply chain planner. Six scenario were studied in detail. For each scenario the
researcher began by observing and documenting relevant events and _ historical
patterns over an extended period (several months). Following extensive discussions
and focus group meetings with the stakeholders the researcher constructed a causal
loop model representing systemic structures - the forces and dynamics that had
influenced the patterns of behaviour in the system.
The constructed Causal Loop Diagrams (CLDs) were then validated through follow
up interviews with key participants. Following the validation of the CLDs,
recommendations were formulated to suggest possible actions for improvements.
Company participants were also asked to brainstorm on possible leverage points for
interventions.
Finally for each case scenario, the researcher probed deeper into the mental model of
the stakeholders (i.e., their assumptions, norms, views). In this paper we present two
case scenarios in relation to performance measure conflicts and trade-offs in
FoodCom’s supply chain. The recommendations derived from the discussions are
illustrated at the end of each scenario to improve or overcome the problem situations.
Figure 1: Four levels of system thinking
Events
Patterns
Systemic structures
Mental models
(Maani and Cavana 2007)
CASE SCENARIOS
From the field notes taken during interviews, informal discussions and document
research, the problem situations and key variables were identified. Stakeholders’
views and thoughts were also sought to construct a conceptual model. The problem
situations were summarised into different scenarios to demonstrate the interaction
between KPIs and the complex issues that FoodCom’s management is currently
facing. The scenarios identified highlight how functional teams interacted within the
supply chain and how KPIs governed by different teams contribute or impede each
other.
Scenario One - Urgent Devanning (Fixes that Fail)
Events
In mid 2006, FoodCom’s Supply Chain (SC) team organised a team meeting to
discuss certain events which were of concern to the managers, as follows:
@ Goods receipting time had increased from five days tumover to nearly eight
days. Goods receipting time can be described as the time it takes for the products
to be devanned (moved out of the container), palletised and receipted into the
system.
@ Out of stock (OOS) products had noticeably increased since March 2006 and
accelerating thereafter.
@ Warehouse and distribution staffs frequently complained about the workload and
stress -
Patterns of behavior
Figure 2 - BOT for Out of Stock in Scenario One
A Urgent devanning
organised to meet the Managers putting
& order for product A more pressure on
joy inwards team
3 OOS continue t
£] OOS increased Urgent devanning incease, ad
4 noticeably since organised to meet the
+; | march and order for product B
© | accelerating after
Interruption to normal
devanning schedule
By
the
>
Time
looking at the trend over time, the OOS situation is increasing continuously and
effort of instructing urgent devann and putting more pressure on inwards team is
making the situation worse. More and more products become OOS and inwards team
is under a lot of stress.
Systematic Structure
Figure 3 - CLD for Urgent Devanning Scenario
CFR
Urgent
oos Devanning
~*~ UU
‘S
Inventory R) Interuption to
ko normal
Devanning
Goods Receipting
Ti
ime s s
Goods .}
Receipting fR) Workload
Target
Staff
Turnove
+S Stress
This scenario is a classic "fixes that fail" archetype (Senge 1990) where well
intentioned actions could cause unintended and often harmful consequences.
Historically, FoodCom has several products that routinely face out of stock (OOS)
situation and leads to a low Case Fill Rate (CFR) - customer service level. In order to
fix this, the supply planners request for warehouse and distribution to organise urgent
devanning, so the OOS product lines can prioritised and receipted into stock to meet
the customer orders and relieving the problem of OOS. However, this quick fix of
urgent devanning results in the side effect of interruption to normal work flow which
have the potential to delay normal scheduled devanning and causing more OOS. In
addition, the double handing of shifting containers around on site leads to an increase
in workload for the inwards team which could have the potential to increase stress and
staff tumover to further delay the goods receipting time. So the side effects undermine
the impact of the intervention and the OOS reverts back to its original condition after
some delay.
Mental Models
Members of the supply chain team recognise that there is an issue that needs to be
addressed immediately, but it is difficult to decide where to start. The supply planners
were under pressure to meet customer orders and reduce OOS; therefore they instruct
the inwards team for urgent devanning and busy fire fighting to catch up ensued
whenever an OOS occurs.
The warehouse inwards team thinks if they just “kept their nose down” and follow the
instructions from the head office by working harder and quicker they could help
relieve the OOS situation. But the harder they try, the bigger the problem become and
the workload just keep increasing. One inwards team leader explained during a phone
conversation:
“On top of further delays in receipting time, the cost associated with this
mess is also increasing. Detention charges (similar to a library book
overdue fine) are also increasing due to delay in returning the containers
back to the freight forwarding companies. Storing a 20 ft refrigerated
container could cost up to $500 extra per day!...We are going out of our
way asking for special favours to arrange special container deliveries
directly from the port of Auckland instead of going through the usual
channel to shorten the lead time. But we are still behind. Do they know
what they are doing?”
This scenario has an impact financially in terms of cost of goods. The management,
seeing the decrease in goods receipting time was not performing efficiently, put more
pressure on the inwards team. Special projects were also set up to investigate the
possibilities of increasing the capacity by adding more people into the inwards team
or having double shifts instead of just single 8 hour shift.
Figure 4 below shows some of the key underlying assumptions held by the staff.
These are shown as ‘thought balloons’ which represent the mental models of the
parties involved.
Figure 4: Mental Model for Scenario One Urgent Devanning
Planning: “Making up for
OOS is critical! We need to
catch up and put through
more urgent devanns.”
SS Urgent
Devanning
‘S
{R) Interruption to
normal
devanning
———- Goods Receipting
Goods Time
receipting target
Management: “we are
getting lots of OOS
and goods receipting
time is decreasing. We
need to put more
pressure on the
inwards team.”
Warehouse inwards team:
“Everything is urgent! We had
to do alot of double handling to
shift the containers around to
do the ones that they want first.
We are working harder and
faster, but we are just getting
further and further behind”
Scenario Two - Poor Case Fill Rate (CFR)
Following from scenario one, the OOS situation also directly impacted on case fill
rate (CFR). This is called ‘service level’ in inventory management. The CFR target
for FoodCom in year 2006 was set at 98%, which means 98% of the time when the
customer orders a product, FoodCom will be able to fulfil the order, and fill the
customer shelves with the desired products. An increase in OOS results in poor CFR
which puts sales and customer services teams under undue pressure to meet sales
targets to keep customers happy. During the monthly consensus meeting in June
between supply chain, sales and finance, several events have been identified in
relation to the CFR.
Events
¢ Pallet count (inventory) in warehouse was down by 2000 pallets
* CFR was performing poorly in 2006, category A was performing at an around
70% on average and a low of 57%
* Number of local truck deliveries increased by 15%
* Cost of failure increased substantially - in the second quarter air freighting
cost alone was over fifty thousand dollars.
Patterns of behavior
As the CFR deteriorates, FoodCom’s customers become more and more impatient and
unhappy. The sales teams struggle to meet their target because some of the products
are either OOS or pass the 4 months shelf life rule. The poor CFR also impacts on
their relationship with the customers and some customers even threaten to de-list the
particular poor performing product if the CFR do not improve.
In FoodCom, the transit time for sea freight imported goods from northem
hemisphere is around 4 to 6 weeks, but due to the poor CFR rate, normally many
airfreights are arranged to shorten the lead time in the hope to solve OOS issues and
satisfy customer's needs. While meeting customer's needs are important, airfreight
charges increase the cost of goods and the cost of failure. Moreover, when the product
finally arrives in NZ, urgent local truck deliveries to the customer need to be arranged
to further shorten the lead time. Some behavior over time graphs are shown in
Figures 5 and 6 below to summarise the situation described:
Figure 5 -BOT for Scenario Two
>
Ccustomer Satisfaction
Cost of Failure
<4
= > ~ > = >
Time Time Time
Pressure to Sale
aed Tl ese,
Figure 6: BOT for Case Fill Rate
Urgent truck deliveries increased
Sales facing pressure by 15%, CFRi da littl
A of products might get fehipeasbaly “aprovecaume
_ de-listed if CFR do not
CSR off target in improve
Sn
2™ Quarter Airfreight - cost of
| / failure increased
/
Customer services team
try to hold back on
orders till a later date
Time
Systemic Structure
The CLD shown in Figure 7 below has highlights the impact of CFR on some key
financial KPIs such as cost of failure, distribution cost and profit. By doing more
airfreights and truck deliveries instead of sea freighting, the product transit time is
shortened in the hope to increase CFR and to satisfy the customers. But the transit
time is shortened at the expense of cost. CFR is only relieved temporarily and
customers are still frustrated which reflects in the low customer satisfactions.
Low customer satisfaction reinforces the communication breakdown between
FoodCom and its customers. As the communication breakdown increases, FoodCom’s
knowledge about their customer plan for promotional activities further decreases the
CFR - hence was forming a reinforcing loop. This is expressed by one territory
manager: “It would be nice if they (customers) have told us about what they were
doing. But low CFR is frustrating our customers, and unhappy customers are less
likely to have open conversations with us.” Moreover, sales team is finding it a lot
harder to negotiate with customers in terms of shelf spacing. “Every time when we try
to negotiate to increase ranging or shelf spacing, customers are hesitant to do so due
to the low CFR. Some even ask us to get CFR back to target before going back to talk
to them.”
10
Figure 7 - CLD for Scenario Two
Truck
Distribution Deliveri
Cost
Cost o
Failure
fen
aN yo OOS
le, on
a ¢ AON
ss s CR) Communication
breakdown
\ customer
Satisfaction
‘Sales
Target
Mental Models
This scenario represented a messy situation that could not be resolved with a quick fix
of airfreight or truck deliveries. The tension between the teams overtime has also
created some blaming culture with undesirable consequences for the organisation.
Sales believe the planning team was not doing an adequate job to ensure there are
enough inventories to meet the customer demand. Finance was also pointing at the
planning team for the same reason which resulted in the increase in cost of failure and
transport costs. The following quotes extracted from a telephone conference involving
all key managers: sales, planning, finance and off-site customer services shed light on
the stakeholders’ mental models:
“We are way off our target CFR of 98%, category A is currently averaging around
70%! I’m constantly getting pondered by unhappy customers about the miss orders
and we got to do something!” Customer services manager
“What happened this quarter? Our airfreight cost has gone through the roof! In
addition, our local delivery charges between North Island and South island have also
increased due to more urgent truck deliveries instead of rail. What’s happening in the
planning team?” Financial controller Supply Chain
“We have over sold in the last quarter and customer A was doing a big promotion on
category A which we did not know about... they didn’t give us enough notice to
respond to the change and depleted all of our south island stock. I understand our
CFR looks horrible but I can only work on what information I have on hand...” said
the supply planner for Category A.
11
“There is no such a thing as over sold - sales have basically under forecast and there
is a communication breakdown.” said the demand planner for Category A.
In response, sales manger defended themselves: “What about the other categories
that are not on promotion? The CFR for those are still off the target. It makes our job
a lot harder out in the trade; some of the customers are even threatening to de-list our
products if our CFR doesn’t improve soon... Our years of relationship with the
customers have been significantly impaired by the poor CFR, planning is making our
job very difficult.” “As for category A, we did not know about the promotion,
customer A has just put the price of the whole range down to treat them as a lost
leader. It will be nice if they have told us about what they were doing. But unhappy
customers are less likely to have open conversations with us.”
“It’s just not good enough, you guys at the front line need to keep us informed, we
can’t keep on doing this. We are spending thousands of dollars rushing around in the
hope to raise our CFR, but our performance seems to be going down even more.” said
the supply planner for Category A.
Following a recommendation to the supply chain manager, the SC team embarked a
daily monitoring of the CFRs. This let to the discovery that a large proportion of poor
performance was the result of customers ordering either deleted items or old product
numbers (run-out lines). When a customer orders a deleted line or an incorrect old
line number, it is also considered as a miss in the CFR report and hence exaggerating
the true CFR figures. To resolve this requires customer services and sales teams to
communicate and collaborate closely.
VALIDATION OF MODELS
After initial data collection and one-on-one interviews, conceptual models qualitative
system dynamics were constructed to develop a picture of the reality. Follow up
meetings with participants were also conducted at this stage to clarify any ambiguous
concepts and problem issues.
The links and relationships between the KPI's were presented visually through the use
of CLDs and the four levels of thinking model. Once the scenarios and CLDs were
validated and finalised, the next step was to bring the participants together to develop
a common vision for taking significant actions on the issues investigated. This helped
to uncover multiple mental models held by the participants.
Next section discusses group sessions where the participants were able to reflect on
and discuss their existing processes and experiences to derive areas requiring change.
Group Discussion Sessions
Several group discussion sessions took place to enable the researcher to present
findings and recommendations back to the organisation. In contrast to initial data
collection interviews, where the participants focused on describing their experience of
12
the issue and the context, this session was to ensure they understand the bigger picture
and explore possible interventions. Group discussion sessions covered three areas:
¢ Introducing the system thinking concepts
¢ Discussion of Scenarios and CLDs identified
« Recommendations for action (intervention strategies)
Overall, all the participants had a positive attitude towards system thinking concepts.
This was demonstrated by their level of enthusiasm and engagement towards the
scenario problems. Some participants were even interested in looking into further
readings about system thinking which was a surprise.
By incorporating multiple stakeholders, participants’ understanding could be extended
and different perspectives could be integrated into a holistic interpretation that
satisfies the different participants. the group discussion happened in several sessions,
each involving the participants that were relevant to the problem scenario.
The group discussions then moved on to discuss the recommendation and to give the
participants a platform for formulating action plans and identifying priority areas for
improvements. Original recommendations of the researcher were validated and if
necessary modified with the rest of the participants in a collaborative approach.
Reflection
At the end of the group discussion session, participants were asked to reflect on what
they had achieved and learnt. The reflections by participants were also taken during
and after the implementation of proposed recommendation.
Participants were asked to evaluate the processes of this research. This information
was valuable to assess the value of system dynamics in enhancing understanding of
the trade-offs between performance measures. and in supporting performance
management decisions.
The focus group like discussion sessions tumed out better than expected. Overall, the
results confirmed that qualitative system dynamics modelling enhanced participant
understanding of the causal relationship between KPIs. Moreover, the process
clarified complex issues involving multiple objectives enabling decision makers to
understand the strength and weaknesses of each approach to make an educated trade-
off decision.
This approach illustrated that KPIs could exhibit contradictory behaviour between one
another, typically across different functional teams, as was the case of increase in
sales which impacted the distribution cost negatively.
some extend demonstrating the Hawthorne effect. However, it is believed the scale
and magnitude of the improvement cannot be simply explained by the extra time and
resource devoted to performance measurement. The performance improvement
observed during the course of this research is the largest increase in FoodCom’s
13
performance as far as the staff can recall. This adds further confidence to the integrity
of this finding.
The debate between marketing and technical and production departments over
strategic product ranging and factory complexity illustrates the trade-off concept
perfectly. On one hand, product ranging is the main drive for sustainable growth and
market share. On the other hand, the increase in factory complexity impacts short
term financial results severely. The performance manager in the factory cynically
referred to as “success is measured by how much we lost, because we are loosing
money on every packet we sell.”
Nevertheless, the performance in cost of production improved as soon as the
‘unnecessary’ packet configurations were deleted. At the same time, the market share
was unaffected as customers simply trade up or down to the next packet configuration
available. This result surprised marketing and changed their stance towards the
strategic ranging aspects.
TEAM LEARNING AND COMMITMENT
Decision making in FoodCom has always been complex and involves several strategic
objectives that are sometimes contradictory. Each function team has different views
of the situation and defines problems differently. The research showed that by sharing
the underlying assumptions and mental models in the personal world, the participants
were able to learn from each other and collaboratively construct a shared perspective
in the social world. The shared perspective in the social world then can be derived
into the technical world to modify system policies, to engage the behaviour of
participants socially and hence change the mindset of people personally. As a result,
conflicts can be minimized between different teams with different objectives and
different K Pls.
The Causal Loop Diagram (CLD) illustrated several key advantages in investigating
the interdependencies amongst KPIs. This provided a clear picture of different
attributes of the problem variables and the interconnectedness amongst them. Cause
and effect, time delays and feedback loops can he illustrated via CLDs and
demonstrate the dynamic behaviour of the system. For example, in Scenario Two, the
KPI case fill rate (CFR) is an important measure for customers. Lower values of CFR
reflect a poor performance of meeting customer deliveries and hence reflecting a
lower customer satisfaction. The CLD constructed showed how this KPI interacts
with the other variables. An increase in delivery costs will lead to an increase in the
CFR and conversely, an increase in communication breakdown resulting from a poor
customer satisfaction which in turn decreases the CFR. Therefore the system thinking
techniques, in particular developing CLDs enhanced participants understanding of the
interdependencies through a holistic view of the system.
Overall, this research created a learning atmosphere to foster shared understanding, as
a result, commitment and direction of the staff in the FoodCom changed. People no
longer cling to “dearly held views” but instead are open to make compromises and to
14
help out other colleagues in different teams. A classic example of this is demonstrated
by Scenario Four - Sales trade spending vs. supply chain contract to clear.
“The key take-home for me from this discussion session will be regarding to
understanding of ‘working as a whole’ concept, that write offs are not SC’s
responsibility alone. Our sales trade spend budget and SC’s write off are really from
the same bucket of funding.” said one key account manager.
Sales team started to change their mindset regarding the balance between trades spend
funding and write off. Through discussion sessions, sales team demonstrated a
sympathetic view towards supply chain by beginning to consider the funding as one
pool of money.
Many participants described organisational learning was the most valuable payoff of
this research. The benefits of system dynamics to explore inter-relationships of KPI
for supporting decision making derived as much from the process as the outcomes of
analysis. To staff members, in particular the research participants of FoodCom, this
research study has influenced the way they think and act. This research is believed to
have improved their shared understanding of the complex issues in particular the
performance trade-offs of the organisation. The participants have also gained more
appreciation of each other's responsible areas and formed a stronger bond with each
others. Some quotes identified from the discussion session have been listed below
regarding to this area:
“I never thought about the problem this way, now I see! I guess I was a little selfish...
I was probably the one that caused all the ciaos in supply chain. I shall arrange more
meetings with supply chain to find out more about what they do.” Key account
manager.
“It’s amazing how I have actually created these problems for myself!” supply planner
“T have got more exposure to other parts of the business especially marketing’ s view
on market share and product ranging. Being away from the head office makes it a lot
harder to know what's going on in other parts of the business. This session is
beneficial, we should get more people involve in similar cross-functional discussions.
It could add a lot of value.” Factory performance manager
GROUP DYNAMICS - ORGANISATIONAL CHANGE
As the researcher gained the trust of the participants and cross functional teams
became more involved both in formal and social settings, closer goal alignment was
achieved both vertically and horizontally.
Although the research disrupted some practices which have been institutionalised at
FoodCom’s for a long time, it is expected that it will result in several positive
influences on the participants and team dynamics. Two interrelated aspects arose from
the action research cycles and were perceived to have contributed to the following
learning and behavioral changes:
15
o Developing and sharing knowledge
o Valuing the big picture
Developing and sharing knowledge
“You cannot have a learning organisation without a shared vision...A
shared vision provides a compass to keep learning on course when
stress develops." — Peter Senge
From the beginning of this research, the behaviour of the participants notably changed
gradually throughout the 18 months of this research. As a direct consequence of a
shared vision, the level of communication between functional groups also increased.
For example, supply chain department has now a stronger involvement with sales and
marketing in product promotions. Sales team begun to take the proactive approach to
check with SC and discuss possible stock level impacts before committing to special
activities in the trade. Sales also developed a trade activity report so SC could have
more visibility in trade activities and adjust demand and forecast accordingly.
By increasing the visibility of functional KPIs, some new opportunities has opened up
for FoodCom. Rather than reinventing the wheel, the teams are now leaming from
each other, sharing information cross functionally using the same report. The
development of a new monthly operations review (MOR) reports had a notable
contribution to the increase in performance of some KPIs particularly in the last
quarter of 2006.
The MOR report contains KPI figures which are used by all functions and each team
is now more aware of what is going on in other areas of the business. For example,
the report has enhanced communication between sales and demand planners and
increased demand plan accuracy. This resulted in Case fill rate to increase to an
average of 94% from around 70%. More importantly, the teams now understand how
their KPIs could affect each others. The systems perspective and successful
collaboration has proven to be gradually helping participants to aim towards feeling as
one team with one common goal.
Valuing the big picture to prevent Sub-optimisation
This research has brought FoodCom one step further by exposing and changing their
underlying mental models to reduce, if not eliminate, destructive competitions
amongst functional teams.
A cross-functional view of performance measurement consolidating several functional
aspects into one holistic picture tends to have greater effectiveness. At the beginning
of the research, sales person A during one informal discussion viewed supply chain
people as whistle blowers:
“SC team is the whistle blower, because they are usually the one that says NO to
marketing’s new product launch ideas, NO to the promotion date that we have
organised with the trade, they are the one that stop all the fun!”
16
But towards the end of the research, when asked to comment on the supply chain
department again the same individual said:
“Tnever thought about an issue this way, now it seems so simple and it is all common
sense. Why didn’t we look [at] the situation this way before? I now understand why
they [supply chain team] are always challenging our ideas. To be honest we never
actually think about whether it is cost effective or whether the inventory level is
enough to go ahead with promotion...[before] it wasn’t really my problem. My only
concern was how we can increase sales from particular activity. I missed the whole
idea of the big picture and was only achieving local optimisation.”
The practice of personal reflection demonstrated a valuable and positive impact that
this research had on staff attitudes. The combined power of action research and
system thinking also contributed to this key success.
Language and Gesture
In addition to the change in mindset, the terminology/language used by the
participants has also changed over the duration of the case study. As the participants
became more familiar with the concepts of system thinking, they have started using its
terminologies and phrases such as: ‘fire fighting; feedback; root cause; goes around in
a loop; and the big picture.” Several of the participants also use hand gestures to draw
a loop in air while explaining their interpretation of scenarios.
“T didn’t really see how I could help the SC department in terms of reducing bad
goods and distribution cost, because after all, my job is demand planning and
concentrating on the DPA and making sure that we meet our target of 75%. But now
through identifying these causal relationships I understand! If the forecasted volume
is too high, inventory goes through the roof; cost of working capital increases.
Moreover, stocks become aged and will need to be written off when the expiry date
hits us. Alternatively, if the forecasted volume is too conservative, we will be out of
stock (started using his hand gesture to draw a loop in the air while explaining) which
hits our CFR. In order to counter attack OOS, we need to arrange urgent devanning,
urgent deliveries by truck to SI, which again increases more OOS. The loop just keeps
on going!”
Research participants are starting to step away from the linear thinking and move
towards closed loop thinking. Hence, systems thinking principles and language can
trigger behavioural change and enhance team learning in complex organisations.
ORGANISATION LEARNING
The purpose of this research was to use system thinking concepts and tools to
investigate the interdependencies amongst KPIs and the causes and consequences of
performance trade-offs in a multinational organisation. The word that most
appropriately describes the broad context for contemporary performance measurement
is uncertainty (Milgate 2004). Organisations are constantly facing an uncertain
17
dynamic world in which they must balance multiple stakeholders, multiple time
horizons and multiple KPIs. In such an environment, performance measurement and
management need to focus on identifying the root causes of performance problems
and understanding the reasons for performance laps. However, even when a
comprehensive analysis of the problems is carried out, the identification of suitable
corrective actions is still considered challenging (Santos, Belton et al. 2002). While
one KPI would indicate high levels of performance, it might have adverse effect on
one or more other areas of performance. Often, the success in one area can only be
achieved at the expense of another, which suggests trade-offs between performance
indicators may be inevitable in complex dynamic systems.
The outcomes of this research can be presented on two levels. First, introducing the
concepts of system thinking and system dynamics modeling (in particularly the CLD)
captured a holistic view of the KPIs and their interdependencies. Understanding of the
cause and effect relationships between indicators and actual performance was
enhanced through the process of identifying events, patterns, systemic structure and
mental models. The case scenarios provided evidence that the system thinking
approach had a substantial impact on improving FoodCom’s KPI performance. Most
research participants agreed that there had been several improvements in the KPI
results in the last quarter of 2006. By combining shared vision and the profound
knowledge derived from personal mental models (identified from interviews) and
collective team learning (through the discussion sessions), system thinking process
enhanced cross functional understanding and cooperation. This research also raised
awareness of the dynamic nature of the performance indicators.
Secondly, by giving the participants the opportunity to engage in the process of
inquiry they increased their knowledge of cross-functional dynamics and changed
their mindset and behaviour - showing more appreciation for the needs and concerns
of other teams which were previously overlooked. This allowed the employees the
opportunity to contribute and create a shared vision. As a result the decisions were
made by consensus and represented balanced perspectives, leading to greater buy-in
and commitments to actions.
Overall, this research had a positive impact on FoodCom’s culture and performance.
It allowed the staff to engage and take ownership of the process and to achieve higher
performance. Although the resultant decision making process might have been
considered as tedious and initially time consuming, the teams realised greater payoffs
down the track. This ‘collective wisdom’ had further flow on effect organisational
behaviour and relationships.
RESEARCH CONTRIBUTION
Although exploratory in nature, this study has important theoretical and practical
implications.
From a theoretical perspective, the study provides empirical evidence that the systems
thinking approach can help managers and staff to gain a holistic view of their
organisations and to understand dynamic interdependencies and trade-offs amongst
18
the performance indicators. The use of the four-level model to frame the research
problem and explain the results made a significant contribution to this research. As
the events were identified and behaviour over time graphs were constructed,
significant insights into the underlying patterns of variables were gained. Together
with the causal loop diagrams, inter-linkages between cause and effect of different
KPI variables were mapped and integrated to demonstrate the systemic structures
underlying performance measurement. The results, insights and proposed
interventions were then communicated back to the stakeholders, which helped them
revise their mental model and facilitated group and organisational learning.
From the managerial practice perspective, this study demonstrated an effective
process to create an open environment for cross-functional communication and
organisational learning. By constructing the CLD models, decision makers could
enhance their understanding of how actions today can affect future performance.
Therefore, decisions could be made proactively and systemically, enabling decision
makers to initiate changes without adverse and unintended consequences. The
proactive rather than reactive management of performance was evident at the case
study organisation.
From the above theoretical and practical outcomes, it is concluded that the use of
system dynamics can catalyse and enhance team dynamics and organisational
learning. It also overcomes the silo mindset and reduces the risk of making erroneous
inferences of dynamic processes imbedded in complex systems.
FoodCom - 16 months on
We revisited the case company 16 months after the initial research and asked three
questions as follow:
Q1: whether the collaboration between the teams had continued?
A: Yes, particularly in the Supply Chain team. There is a strong consensus amongst
the entire SC division that communication and cross-functional collaboration had
become the key area of focus for business success. The CEO commented that SC
(and in particular the planners) is no longer just the service provider as they are now
actively involved in the commercial part of the business (sales). The more SC
understands the sales promotional plans, shelf strategy, pricing strategy...etc, the more
they can accommodate and even challenge the commercial part of the business.
Q2: whether the SC performance improvements have been sustained?
A: The main SC KPI results (DPA, CFR, Stock Cover, bad goods, etc) have all shown
consistent improvement. CFR, for example reached 97.5% in 2007. The
improvements are in large part due to enhanced dialogue, awareness of potential KPI
trade-offs and a deeper understanding of what other departments do and how their
actions affect each other.
Q3: whether the employees have continued to utilise systems thinking concepts to
maintain a culture of continuous improvement and learning?
19
A: Since the introduction of Systems Thinking, FoodCom has initiated a “One
Number principle” - a holistic view of business where cross functional stakeholders
collaborate together and share responsibility to achieve a single company target. This
is in sharp contrast to the previous mode of operations whereby sales set its own
target, SC forecast another figure, and finance would budget for another number
which resulted in endemic dysfunctional behavior. Although this is still work in
progress, it is believed that the benefits from adopting a systemic performance
management is profound and enduring.
REFERENCES
Boume, M., J. Mills, et al. (2000). "Designing, implementing and updating
performance measurement systems." International Journal of Operations &
Production Management 20(7): 754-771.
Castellano, J. F., S. Young, et al. (2004). "The seven fatal flaws of Performance
Measurement." The CPA Journal 74(6): 32.
Ferdows, K. and A. De Meyer (1990). "Lasting improvements in manufacturing
performance: in search of a new theory." Joumal of Operations Management 9(2):
168-89.
Forrester, J. (1994). Learning through system dynamics as preparation for the 21st
century. System thinking and dynamic modeling conference for K-12 Education,
Concord, USA.
Johnston, Brignall, et al. (2002). ""Good enough" performance measurement: a trade-
off between activity and action." Journal of Operational Research Society
53: 256-262.
Kaplan, R. S. and D. P. Norton (1992). "The Balanced Scorecard Measures That
Drive Performance." Harvard Business Review 70(1): 9.
Lynch, R. L. and K. F. Cross (1991). Measure Up: Yardsticks for Continuous
improvement. Blackwell, Cambridge.
Maani, K. E. and R. Y. Cavana (2007). Systems thinking and modelling:
Understanding change and complexity. Auckland, Pearson Education New Zealand,
24 edition.
Milgate, M. A. (2004). Transforming Corporate Performance: Measuring and
Managing the Drivers of Business Success, Praeger Publishers.
Neely, A., C. Admans, et al. (2002). The Performance Prism - The scorecard for
measuring and managing business success, Pearson Education Limited.
20
Neely, A., K. M. P, et al. (1999). "The forces that shape orgnisational performance
measurement systems: An interdisciplinary review." International Journal of
Production Economics 60: 53-61.
Santos, Belton, et al. (2002). "Adding value to performance measurement by using
system dynamics and multicriteria analysis." International Journal of Operations &
Production Management 22(11): 1246.
Senge, P. (1990). The fifth Discipline The art_& Practice of The leaming
Organisation. London, Century Business.
Silveira, G. D. and N. Slack (2001). "Exploring the trade-off Concept." Intemational
Journal of Operations & Production Management 21(7): 949-964.
Todd, D. P. (2000). A "Dynamic" balanced socrecard. Management Science and
Information Systems. Auckland, University of Auckland.
Yin, R. K. (2003). Case study research Design and Methods, Sage Pblications.
Youngblood, A. D. (2003). The Use of Multi-Attribute Utility Theory to Address
Trade-Offs for the Balanced Scorecard. Department of Industrial Engineering
Fayetteville, University of Arkansas: 222.
21
