Analysing the dynamics of mental models using causal loop diagrams

Abstract

Substantial evidence suggests that managerial mental models play an important role in firm performance. Yet managerial mental models are not static but dynamic. This research investigates the creation and evolution of mental models over time and how this dynamic process influences strategic choice and firm performance. We adopt the causal loop diagramming method, with in-depth case analysis over a period of 12 years, as the primary investigatory approach. Our research contributes to knowledge by identifying the shared mental model of the top management team, represented in the causal loop diagrams, for each stage of the company’s development. Our findings suggest that the dynamics of managerial mental models explains the changes in firm performance over time.

JEL classification: L2

Keywords

causal loop diagrams mental model performance change

1. Introduction

Substantial evidence shows that managerial mental models play an important role in firm performance (Barr et al., 1992; Gary et al., 2012; Reger and Palmer, 1996; Walsh, 1995). Managerial mental models are ‘the simplified knowledge structures or cognitive representations of how the business environment works’ (Gary and Wood, 2011: 569). Management scholars show keen interest in how mental models affect strategic choices and firm performance, and investigate their relationships from different perspectives, such as cognitive science and strategic management (Barr et al., 1992; Tripsas and Gavetti, 2000). Importantly, mental models are not static but dynamic (Walsh, 1995). When managers receive new information, they retrieve that information using their current mental models in order to determine strategic actions. Based on the results of these actions, managers may update their beliefs and assumptions accordingly (Vandenbosch and Higgins, 1996). However, few studies have examined the creation and evolution of mental models over time and how this process influences strategic choices and firm performance.

This study takes a more dynamic approach, by examining how managerial mental models form and change in an organization and how the dynamics of mental models affects firm performance over time. Prior studies have largely focused on the sources of mental models in large, well-established firms (Tripsas and Gavetti, 2000; Yang et al., 2016) and have paid little attention to the origin of mental models of a start-up company and the evolution of these mental models. As a result, there is a paucity of research on the dynamics of mental models and performance consequences over time. The limited understanding of the creation and evaluation of managerial mental models is concerning, given the salient role of mental models in affecting firm performance (Barr et al., 1992; Gary et al., 2012; Reger and Palmer, 1996; Walsh, 1995).

Our empirical evidence is based on a case study. We take the conceptual lens of dynamic decision-making and use the causal loop diagramming (CLD) approach (Perlow et al., 2002; Repenning and Sterman, 2002). CLD is employed because there are multiple and interacting processes, time delays and non-linear effects (e.g. feedback loops) involved in the case, and CLD is regarded as an appropriate investigatory approach for such situations (Sterman, 2000). The case company, Wonderland,¹ is a leading surface-mount technology (SMT) manufacturer in the printed circuit industry. The performance of the company has varied significantly despite the steadily increasing global economy. Led by Dr. Wood, an entrepreneur and the firm’s CEO, Wonderland successfully and dramatically accumulated a sales record from US$5.43 million in 1995 to US$140.10 million in 2006. However, the firm’s path to growing its business performance was not always a smooth one. Between 2000 and 2001, sales revenue slumped from US$20.25 million to US$8.20 million, and this poor performance lasted for more than 3 years. An intriguing question thus arises: under the leadership and management of the same entrepreneur and the same management team, why did this firm exhibit such a high level of changes in its performance level (either low to high or high to low) during this 12-year period? Organizational performance has been regarded as a significant phenomenon of interest in the wide range of studies in strategic management research (see, for example, (Lee et al., 2009; Shea et al., 2012; Welch, 2003). Thus our research question is particularly significant, considering that, by the end of 2006, the same firm had achieved an all-time-high performance record of US$140.10 million even as the business environment changed dramatically.

In-depth analysis of the case company provided us with an unusual opportunity to examine the dynamic process of mental models. We identified the shared mental model of the top management team, represented in the causal loop diagrams, for each stage of the company’s development. By doing so, our research extends understanding in the mental models literature by providing more nuanced insights into the changes of mental models over time and how such changes consequently affect a firm’s strategic choices and performance. Our contribution focuses on the changes in mental models of the same management team (Thomas, 1988), whereas prior studies have examined the ways in which different mental models from different leaders influence strategic decisions and organizational performance (Gavetti, 2005; Kaplan and Tripsas, 2008; Reger and Huff, 1993). Our research advances knowledge by examining the effect of the change in mental models of the same managers on strategic decisions and organizational performance.

2. Managerial mental models and firm performance

Since managers have limited cognitive capabilities, they rely on mental models to make sense of the world (Bandura, 1988; Simon, 1955; Tripsas and Gavetti, 2000). Managerial mental models are organized knowledge structures or simplified cognitive representations that managers use to understand business and its environment (Gary and Wood, 2011). Prior studies have also used terms such as cognitive maps, causal maps, dominant logic, frames and belief systems (Hodgkinson et al., 2004; Knight et al., 1999; Simon, 1991) in analogy to mental models. Managers conceptualize their business and build cause-and-effect understandings about the environment based on their mental models (Barr et al., 1992; Prahalad and Bettis, 1986). Porac et al. (1989) propose two types of beliefs that construct mental models: (1) beliefs about the identity of the firm, its competitors, suppliers and customers and (2) beliefs about how to succeed (i.e. to change or not) in a competitive and changing environment. Some studies have attested to the importance of these sets of beliefs individually. For example, Boulding et al. (1994) conclude that ‘customers, competitors and managers themselves’ are three important antecedents in shaping managers’ mental models. Karakaya and Yannopoulos (2010) observe that, when the technological environment changes, managerial mental models of the business environment shape local incumbents’ responses to competitive threats from global markets. Nevertheless, there remains limited understanding about how these two sets of factors evolve once they are perceived (e.g. an entrepreneur’s beliefs) in combination as the competitive environment changes.

Extensive research evidence shows that managerial mental models are heterogeneous (Barr et al., 1992; Martignoni et al., 2016; Porac et al., 1989; Tripsas and Gavetti, 2000; Walsh, 1995). The differences in managerial mental models can result from managers’ own individual experience and learning capabilities (Gary and Wood, 2011), their firms’ history (Benner and Tripsas, 2012) or even national contexts (Fassin et al., 2015). Martignoni et al. (2016) summarize two different factors that account for the variance in managerial mental models: ‘cognitive types’ that are associated with managers’ innate properties and a ‘cognitive approach’ that is related to more consciously chosen elements. These two factors form ‘cognitive styles’ that managers develop in their mental models.

Prior research has suggested that mental models have great impact on strategic choices and firm performance (Csaszar and Levinthal, 2015; Gary and Wood, 2011; Gavetti, 2005). For example, Bagdasarov et al. (2016) find that greater knowledge, as indicated by mental model complexity, can lead to more ethical decision-making through the mechanism of sense-making. Mental models also interact with search processes and affect the quality of strategies that managers can discover (Csaszar and Levinthal, 2015; Gavetti and Levinthal, 2000). It is widely accepted that more accurate mental models about the key causal relationships in the environment lead to better firm performance (Denrell, 2004; Gary and Wood, 2011). Recent research provides a more nuanced, contingent view of the performance effect of mental models and has proposed the notion of ‘cognitive fit’ (e.g. Martignoni et al., 2016). That is, firm performance is dependent on the degree of cognitive fit between managers’ mental models and environmental properties.

Most extant studies focus on the effect of mental models on strategic choice and performance from a static perspective. For example, Roxas and Coetzer (2012) examine how managerial mental models about the natural environment interact with the institutional environment to affect strategic orientation towards environmental sustainability. Another example is Sterman et al. (2007), who investigate the consequences of firm growth strategy in a competitive environment and reveal that growth strategies can lead a firm into financial crisis if the managers are unaware of bounded rationality during their decision-making processes. Oliva and Sterman (2001) find that, when management practice interacts with the characteristics of services (e.g. intangibility, labour intensity and inseparability), it erodes service standards and lowers the service quality for the whole industry. In short, these studies primarily focus on a specific set of mental models rather than a holistic and dynamic view: that is, how mental models are continually shaped from various sources and how this process of change influences decision-makers’ strategic decision processes and performance changes. Significantly, mental models may not be static. When managers receive new information, they may retrieve information through their mental models, modify their mental models via a feedback loop system (e.g. actions and performance) and accommodate the new information using adjusted mental models (Vandenbosch and Higgins, 1996). Nevertheless, we acknowledge that prior studies also suggest that environmental changes may not necessarily lead to the update of mental models, because such changes may not be significant to the existing models (Barr et al., 1992). Also, the update of the mental model is more likely to occur up to the point where the pressure to change exceeds cognitive inertia (Barr et al., 1992; Ginsberg, 1988). Hall (1984) makes some initial steps in this area by proposing a process model of organizational policymaking to explain the evaluation of mental models and their performance impact over time. Barr et al. (1992) find that cognitive change that accords with changes in the environment is critical to organizational renewal. Conversely, the failure to adapt managerial cognition to changing contexts may lead to organizational inertia and deteriorating performance (Cho and Hambrik, 2006; Tripsas and Gavetti, 2000). Despite some research progress, the paucity of research in this area and the challenge to obtain data of sufficient richness over an extended period impose both theoretical and empirical challenges to understanding the dynamics of mental models. Our study sheds light on this research gap.

3. Methods

We use causal loop diagrams to demonstrate our empirical findings from the interviews and discussions in workshops. The CLD approach is often utilized to show the causal relationships between a number of interacting variables (Morecroft and Sterman, 1994; Perlow et al., 2002). We contend that CLD is an especially useful method for our study, because it is widely applied in situations with interacting processes, time delays and non-linear effects (Davis et al., 2007). To employ this method, we review the ‘story or narrative’ of the case company from the interviews and discussions in workshops, identify the main variables that emerge from the data and build the causal linkages between variables. The resulting causal loop model is grounded in our data and generates insights about how different managerial mental models are interacted when some variables are controlled.

3.1. Empirical data and data collection

The researchers for this study conducted unstructured, open-ended interviews in order to generate rich data, following Patton’s (1980) suggestion. During the interviews, reflective conversations were conducted to reveal the managerial mental models of the focal company. To ensure internal validity and the accuracy, trustworthiness and coherence of information (Lincoln and Guba, 1985), multiple interviews were undertaken for each division of the company. Interviewees largely comprised management team members, including the general manager, senior managers and division managers. A small group of professional employees (e.g. senior engineers and senior employees) was included to provide a richer understanding. The average working experience for interviewees was more than 10 years. We organized two interviews each week. The weekly interviews lasted for 6 months from February to July in 2007 in Kaohsiung and Shenzhen. In total, we conducted 48 interviews. Each interview lasted for 20–30 minutes via face-to-face or telephone conversation. The empirical evidence is also supplemented by archival and historical data collected from Wonderland’s database and the government’s public information database. During the data collection process, one consulting team was involved in conducting systems-thinking workshops and group-modelling sessions regarding mental models with senior management in a computer laboratory at a university in Kaohsiung. The systems-thinking workshops were conducted fortnightly from April to July in 2007. In total, eight systems-thinking workshops were conducted. Thanks to these workshops, the researchers were able to develop the CLDs and are confident that the CLDs of the company have been validated internally by the participants in those workshops.

3.2. Data analysis

We took several steps suggested in prior research to convert the interview data to CLDs (Sterman, 2000). In the initial stage of the data collection process, we asked participants/interviewees two questions: How do you describe the historical performance patterns (e.g. profits, revenue) back to the identified time period? What were the key factors driving such performance pattern? Based on their responses, we drew some basic CLDs that demonstrated the historical events and case company situations. Most importantly, several recurring themes emerged from the interview data (e.g. product quality, adoption rate, material price, delivery delay, turnover rate) and they were included as the key variables in the CLDs. In the second step, interview data and the workshop discussions were further analysed to identify the causal links between themes. These themes and links form the basis for the CLDs that we developed. In this process, there is some divergence regarding the themes and links. To address the divergence, in the third stage we developed a representation of business as a system with collaborative effort.

We used these basic CLDs as the starting point for the systems-thinking workshops that followed. In the workshops, we sought to complete these CLDs with the participants by asking them to verify the causal links together. The process began with the selection of patterns of interest from the initial CLDs and data analysis and then continued with the iterative developments of CLDs. The variables and causal links in CLDs captured in the discussion in the workshops formed the feedback structure of the company. Later that updated feedback structure generated the dynamics of the system that the focal company actually occupied.

Next, we further developed the CLDs by describing particular relationships between variables to explain the changes in performance over time. We then integrated these unified diagrams into a single set of feedback processes which was capable of explaining the multiple behavioural patterns of observed historical events. In this process, we often returned to the data to check for and resolve any contradictions. In the workshops, we regularly reviewed our results with participants from the focal company and members in the research team. All of them provided insightful additions and clarifications. The confirmatory evaluations from the interviewees provide validation that the CLDs had captured an accurate representation of managerial mental models over time. Based on this back-and-forth procedure between the research team and the interviewees, we are confident that the mental model represented by the CLDs is a company-wide mental model interpreted similarly by individuals within the company.

As our model emerged, we reviewed each link in the CLD to assess whether the observed relationship was supported by the field data or existing studies in the literature. This step helped us to ensure that the model was grounded in the collected data and was consistent with the primary principles of the decision-making and strategic management literature.

4. The case company

Wonderland Company, founded in 1989, is a leading SMT manufacturer in the printed circuit industry. Over a 12-year period, Wonderland saw its sales record rise dramatically from US$5.43 million (1995) to US$140.10 million (2006). Throughout its history, the firm’s strong R&D team enabled the company to keep pace with its competitors as a primary competitive advantage. Wonderland’s applications included add-on cards (peripherals), motherboards, telecom products and, eventually, radio base stations, mobile phones, digital cameras and LCD monitors/TVs/notebooks, as well as other digital and entertainment products. This study examines the history of the case company from 1995 to 2006, a period of time that can be divided into three stages: the collectivity stage, the formalization stage and the elaboration stage.

4.1. Stage 1: collaborative stage (1995–1998)

At the collaborative stage, the company focused on survival in a competitive environment. The entrepreneur believed that only products of superior quality could capture potential customers. As the founder CEO, Dr. Wood, highlighted:

We believe the quality of our products is the best way to convince a prospective customer. Quality is always our first priority. (Interview 4 with the CEO)

Because of this belief, the company made every effort to obtain quality certificates. As is shown in the company’s historical report, by the end of 1997 the company had been awarded a QS9002 certificate; it had also received quality approval issued by Compaq and Motorola, which enabled Wonderland to become a business partner of these two companies. In addition, these well-known certificates helped the company update its technologies and earned it a good reputation and more business (Chow-Chua et al., 2003; Terziovski et al., 1997). At this stage, the company’s size grew rapidly, both in number of employees and volume of equipment. One senior sales manager recalled:

During that time period, the number of production lines expanded from three to six, and the company’s hierarchical organizational levels expanded from three to five. (Interview 8 with Senior Sales Manager 1)

Based on the interviews and discussions in the systems-thinking workshops, we suggested five feedback loops at the first stage (three balancing feedback loops and two reinforcing feedback loops) that determined the growth of the business and the consequential firm performance at this stage. These causal loop diagrams illustrate the perceived causal relationships of the managers’ mental models. Reinforcing feedback loop 1 (R1) refers to the customer adoption process. The customer’s decision to adopt (or not) the products of the focal company is conditional on two factors: relative price attractiveness and relative product quality attractiveness. As the competitors decrease their product price or improve their product quality, the relative attractiveness of focal firm’s price or quality will decrease, respectively. Consequently, customer adoption rate decreases. We contend that the latter is in some respects more important than the former. One senior sales manager explained this:

Customers use a product’s price to determine if the product is affordable to them. Additionally, customers also appear to use a product’s quality as an important evaluation factor to make their buy-no-buy decision. Although these two factors are both crucial to a customer, my experience also shows that, when a customer has some uncertainty concerning a product’s quality, they tend not to take the risk. (Interview 30 with Senior Sales Manager 2)

Note that we include price and product quality in the CLD as exogenous variables (see Figure 1), because the CEO and senior management team regarded them as the policy decision variables. The CEO purposely determined fixed goals for both price and product quality to effectively attract potential customers. While these goals tend to be stable over time, we also acknowledge the possibility that the price may slightly change due to the competitors’ price and product quality. Thus, we also include those two variables in the model. However, the main decision factor in the focal company remains the CEO’s and TMT’s decisions. In addition, we include structure development in the CLD as another exogenous variable (see Figure 4). This variable originates from a formal policy determined by senior management on a specific purpose of communication efficiency. We introduce this policy in Stages 2 and 3.

Customer orders are positively associated with customer adoption rate. Specifically, when customer orders increase, the company needs more sales groups to deal with the increased orders. The increased sales groups provide faster and better customer relationships through marketing and business communication. Sales capability improves accordingly. As sales capability of the company increases, customer adoption rate again increases commensurately. As was shown in the company’s sales report, initially there were four sales groups. As customer orders started to increase, the number of sales groups increased to six in Stage 1.

Reinforcing feedback loop 2 (R2) refers to profits and growth. This is the positive loop that drives the company to grow over time as customer orders increase, as predicted by reinforcing feedback loop 1 (R1). When customer orders increase, profits also increase. The increase in profit may relieve the firm’s financial pressure and give senior management the opportunity to invest in company’s assets. The company then has sufficient equipment to provide reliable and available capacity to meet customer requirements and satisfy their orders. Consequently, customer orders increase. Thus, we contend that R1 and R2 drive the growth of the firm and increase firm performance.

Balancing feedback 1 (B1) refers to firm capacity. When customer orders increase, requested materials volume for production increases accordingly. Thus, cost of materials increases along with the requested materials volume. Company debt, therefore, is increased significantly and the company has less cash on hand to make investment in replacing old equipment. Consequently, available capacity decreases. This is crucial, because customers are more likely to place new orders when they perceive that their suppliers have sufficient and stable capacity. Otherwise customers are likely to seek alternative suppliers.

Another balancing feedback loop that is strongly related to B1 is B2 — the Account Receivable Loop. Financial pressure arising from increased orders leads managers to reduce the actual account receivable period (Kroes and Manikas, 2014; Moss and Stine, 1993). The account receivable period is critical to customers, given the financial pressure on their own resources. Thus, when the account receivable period decreases, customers experience financial pressure. As a result, customers place fewer orders with the focal company. The focal firm then requires less requested materials volume, and this therefore lessens cost of materials. Consequently, profit increases and financial pressure is relieved. However, the cost of this management decision is the loss of customer orders. This insight was captured by one senior finance manager in a workshop:

The normal account receivable period is eight weeks. But it is conditional on the relationship between supplier and buyer. In a buyer market, for example, if the buyer is a Japanese firm, the account receivable period is relatively stable, given our company has limited negotiation power with them. However, if it is a supplier market, the account receivable period then depends on the financial pressure of our company. If the pressure is high, we would hope to reduce the account receivable period to relieve our pressure. Thus customers would sense the pressure, because we shift the burden to them, and they were forced to find other suppliers who could provide a longer time period of account receivable. So our cash pressure would kill our orders gradually. (Interview 19 with Senior Finance Manager 1)

Financial pressure also stimulates one reinforcing feedback loop (R3)— the Account Payable Loop.

Financial pressure drives managers to increase the actual account payable period. A sales manager highlighted this:

Recently, we just increased the period of account payable from four weeks to eight weeks. When suffering from financial pressure, we would increase the period of account payable. In doing so, the speed of cash outflow becomes slow. Senior management believes that a good use of financial control capability would be able to let the company make higher profits with lower capital. (Interview 42 with Senior Sales Manager 3)

Consequently, suppliers of the focal company start to charge higher materials prices, and the cost of materials increases accordingly. The profits of the focal firm then drop. In turn, financial pressure is strengthened in such a reinforcing feedback loop.

Balancing feedback loop 3 (B3) refers to the staff adjustment process. When orders steadily increase, the focal firm needs more employees due to concern for productivity. Labour costs to the focal firm include fixed costs and variable costs. Senior management can make the strategic decision of hiring more full-time or part-time employees (i.e. Decision on Hiring Full-time Employees in Figure 1). Hiring full-time employees is associated with fixed cost; hiring part-time employees is associated with variable cost. Both full-time and part-time employees increase labour costs. Thus profits decrease as labour costs rise. Consequently, financial pressure is reinforced. When financial pressure increases, investment in fixed assets diminishes, the amount of equipment reduces and the firm’s available capacity decreases. In turn, customer orders suffer from this limited capacity. Simultaneously, financial pressure triggers B2 and R3. Thus the staff adjustment loop may eventually limit the expansion of the focal firm and opportunities in the future. Figure 1 shows the causal loops articulated above.

We observe that, interestingly, the focal firm’s sales revenue and profits did not stay flat or drop as suggested under the balancing feedback loops, B2 and B3, or under the reinforcing feedback loop, R3. Rather, the historical pattern implies the opposite trend. Figure 2 shows the revenue pattern from 1995 to 1998. Revenue in fact increased from US$5.56 million to US$12.13 million in 3 years. Figure 3 shows the profit pattern of the focal company: profit increased from US$0.16 million to US$2.26 million in 3 years.

In the interviews and discussion in the workshops, we found that the entrepreneur CEO’s beliefs on quality drove the decision-making and helped alleviate the negative effects of financial pressure on firm performance – the core issue in the early stage (Stage 1). That is, loop R1 dominated in Stage 1 and drove firm growth. Specifically, the CEO believed that quality was the key criterion for making company products attractive. He acknowledged that the depreciation speed was faster in electrical products such as printed circuit boards (PCB). When products were delivered to the firm’s customers and there was a quality issue, customers did not ask for a replacement but for a direct refund. This was because the PCB was the main container of other components. Once the main container developed problems, all other electrical components installed on the board from other suppliers were damaged. Hence, when customers sought repayment, they not only asked for the costs of the PCB but also for those of other components installed on it. One senior manager in the executive office recalled that ‘the repayment would be sometimes ten times higher than the PCB board itself’. The CEO therefore looked closely at the quality of the PCB. And, given this insistence, customers were happy to adopt the focal firm’s products, even though the price was relatively higher than that of other suppliers.

In addition, when customer orders steadily increased, the CEO was inclined to hire full-time rather than part-time workers (see Decision on Hiring Full-time Employees in Figure 1). A finance manager commented:

In general, the cost of one part-time employee per hour is 80% higher than one full-time employee. Therefore, to reduce the financial pressure caused from the increased labor costs [in the B3 loop], we preferred to hire more full-time employees. (Interview 19 with Senior Finance Manager 1)

Most importantly, one significant trade-off between full-time and part-time employees is that full-time employees become essentially a fixed cost once they are hired, while part-time employees are always variable costs. As a result of a policy to hire more full-time employees, the case company had better control over cost and consequently significant increased product profits. Financial pressure was relieved accordingly. Of course, in the case company, the investment decision on hiring full-time instead of part-time employees was very risky. It is commonly believed that a full-time employee incurs higher employment costs than short- and fixed-term employees (e.g. basic salary, employment taxes and benefits). Thus, this was an ambitious decision that relied on the entrepreneurial CEO’s vision and mission on product quality. Full-time employees received better training programmes and job security, and therefore product quality was enhanced.

4.2. Stage 2: formalization stage (1998–2002)

At the formalization stage, the company started to develop formalized policies, rules and regulations, and the communication between departments was likewise formalized. The launch of this stage was driven by the senior management’s persistent desire for product quality, their emphasis on efficiency and the changes in the business environment that imposed pressure on cost control. At this stage, the CEO and senior management continued to value the importance of product quality. A vice president recalled:

The cornerstone of our firm’s success is offering a high-quality product. If a popular item sells for a very low price, and it is poor quality, you will lose customers very quickly. The key is to provide that which has great value, as this will increase repeat patronage and word-of-mouth advertising, building a good reputation for your firm’s reputation. (Interview 18 with the Vice President)

Given the significant increase in the number of customers and orders in the first stage (1995–1998), the firm had to expand in size to cope with the corresponding increase in customers. Firm expansion forced the company to hire an increasing number of new employees. The senior management soon recognized the quality problem caused by the novices. One senior engineer recalled:

It took the new hires 8–10 months to make them become productive. Also, it would cost the experienced employees a significant amount of time to train new hires and, even worse, they had to work overtime. (Interview 11 with Senior Engineer 1)

This issue was implied by the balancing feedback loop 4 (B4)—Delivery Delay Loop, which dominated in Stage 2 and affected consequent firm performance. Increased orders made the situation of working overtime severe. A vice president recalled:

We have three shifts per day. Each shift is eight hours long. Normally, working overtime is limited to two hours. If we ask employees to work more than 10 hours per shift, you can see the efficiency and quality drop quickly. (Interview 18 with the Vice President)

In such a situation experienced employees will choose to leave. Thus, turnover rate of experienced employees increases accordingly. Note that novices need a substantial amount of time to become productive. Thus the work quality decreases due to the significant loss of experienced employees. Consequently, the product rejection rate rises, so that delivery delay becomes longer. Eventually, the customer adoption rate suffers. This quality control ‘fade’ was recognized by the company’s senior management at the time. The CEO therefore determined to formalize working procedures to ensure efficiency, quality of employee work and product quality at the same time. From 1998 to 2002 many systems were introduced to the focal firm. The CEO recalled:

My concern is the quality. I am thinking: how can our firm provide better quality products to our customers? As the firm size expanded so quickly, the pressure on experienced employees increased enormously. When those experienced employees left, you should expect the product quality to drop accordingly. To address this issue, I think we need to rely on the system and structure rather than employees, especially as the turnover rate is climbing. Therefore, we started to design a formal communication system within the firm. I think we need to ask our employees to follow SOPs [Standard Operation Procedures]. All the communication between employees needs to be formally recorded. The most important change in our firm is the adoption of an ERP [Enterprise Resource Planning] system. … I understand the potential risk and cost associated, but if we do not do it now, it may be too late in the near future. (Interview 4 with the CEO)

Nevertheless, as the business environment changed, the pressure from customers to lower costs became severe. As one senior sales manager said:

Our competitors are making good progress in their product quality. Their technology level becomes increasingly closer to ours. To increase market share, the competitors constantly reduce their price to attract more customers. As a result, the competition goes up and the customers have a higher bargaining power over price. (Interview 8 with Senior Sales Manager 1)

The effects of the improvement in competitors’ product quality and price are also captured in Figure 4: when competitors’ product quality increases or product price decreases, the relative product quality and price attractiveness decreases accordingly, and hence the customer adoption rate suffers.

The changes in the business environment forced all firms in this industry, including the focal company, to seek alternatives to minimize costs: for example, the company started to outsource to cut costs. To better manage the outsourcing partners, senior management believed that the firm had to speed up its formalization process in order to cooperate with different factories and suppliers. For example, the company adopted an enterprise resource planning (ERP) system to improve the speed of formal communication.

Given the ongoing changes to the focal company, such as the ERP and organic structure, significant time and resources, such as employee training, were involved in this process to implement the changes. Due to the necessary resources involved, sales revenue suffered at this stage (see Figure 2). In the period between 1998 and 2002, revenue remained stable from 1998 to 2000, but dropped significantly after 2000. Profit dropped significantly after 2000 and was below zero after 2001 (−US$4000 in 2001 and −US$0.47 million in 2002). The revenue and profit patterns demonstrated the costs of the adoption of those formalization systems and the change in organizational structure.

At this stage, we observe the changes in mental model together with the interaction between the beliefs of the entrepreneurs and the dynamics of the business environment. The entrepreneur CEO was keen to improve product quality to maintain product attractiveness to customers. The change in the business environment forced companies in the industry to control costs and pay extra attention to product price. The increase in customer orders in the prior stage also stimulated the B4 feedback loop, the Delivery Delay Loop. The joint effect of those three factors on the customer adoption rate (delivery delay, price and product quality) led the senior management team to carefully investigate the business environment to move the focal firm into the ‘formalization stage’. Although the transition to such a stage was not cost-free, the senior management team and the CEO expected to fundamentally change the operations of the firm by formal communication and organic structures (see Structure Development in Figure 4). These changes were expected to significantly improve communication efficiency in the focal firm, and consequently decrease delivery delay. The top management team expected that these changes would shape the firm in the right direction in the long term and the firm would bear the short-term pain as a trade-off.

4.3. Stage 3: rapid expansion stage (2002–2006)

At this stage, the challenges from the business environment became more severe than in the previous two stages. Given the high level of competition within the industry, customers recognized that it was a ‘customer-take-control’ market (Hagel and Rayport, 1997). Hence, customers started to request more from their suppliers, including lower costs, shortened delivery delay and higher product quality. One sales manager highlighted the challenge in a workshop:

We understood that the customer is always right. We listened to what customers or clients said about the problems of our products and their concerns. We tried hard to satisfy the various needs of our customers. The chief issue with the various needs is that we need sufficient stock in storage for various products (due to the various needs). (Interview 33 with Senior Sales Manager 2)

Given the changes in the business environment, as orders increased stock inventory decreased very rapidly compared to prior stages. This was because customers’ needs were much more diversified than previously. This triggered the balancing feedback loop (B5), the Stock Inventory Adjustment Loop, which was critical to firm growth in Stage 3. Due to significant decrease in stock inventory, the stock available on a weekly basis diminished (i.e. Weeks of Inventory). Thus delivery delay increased. This then decreased customer adoption rate of the firm’s products and hence customer orders suffered. The balancing feedback loop 5 is shown in Figure 4. Furthermore, B5, the Stock Inventory Adjustment Loop, also reinforced the effect of the delivery delay loop, B4, due to increases in delivery delay.

Senior management acknowledged this environmental change and had become prepared for this challenge since the previous formalization stage. The mental model of the CEO and senior management team had significantly changed from quality focus to considering all key factors at the same time (delivery delay, price and product quality). As one senior sales manager recalled:

Our competitors achieved a similar level of product quality to ours. As the level of competition increases, customers become sensitive to product price. At the same time, because customers are not willing to take any risk on their own production lines, the requirement on low delivery time becomes higher as well. (Interview 42 with Senior Sales Manager 3)

The senior management fully understood that their company needed to make changes in order to keep their customers. Recall that, in the formalization stage, the company started to adopt various systems to improve communication formalization and change the organizational structure from bureaucratic to organic. Investment in these two organizational changes was not in vain. When the newly adopted system was widely accepted, production costs were significantly reduced due to the B4 loop’s effect. The CEO said:

I believe that, in order to successfully meet customers’ needs, we need to provide low-cost, high-quality products with prompt delivery. To accomplish these stretch goals, we relied on precise coordination between departments within our company and smooth collaboration between our company and our suppliers to ensure prompt delivery. The ERP system and organizational structure change made these stretch goals achievable. (Interview 4 with the CEO)

Due to the nature of the organic structure, cross-function teamwork became the normal practice in the company. In addition, ERP systems enabled decision-makers to make timely, effective and informed decisions, to increase productivity and to grow profitability. ERP successfully integrated different (software) platforms designed to consolidate information across the entire company – including financials, sales, customer relationship management (CRM), stock inventory and operations. Suppliers were also requested to adopt the same data exchange standard to communicate with the ERP system. This further enhanced the speed of formal communication.

At this stage, too, LCD-related product lines became a main production focus for the company, as this potential market was expected to grow exponentially. The company sensed that, if it could not grow quickly and dominate the market, it would be very likely to lose its initial competitive advantage. Therefore, the company simultaneously expanded its production to a few low-cost cities in mainland China and sought cooperation with additional outsourcing companies to seek further low-cost opportunities. At this stage, competition within the market quickly intensified as more and more competitors joined the low-cost game. The senior management team realized that there was no effective way to stop the competition game. However, given the assistance that structural change and the ERP-related system adoption provided, the company, under the leadership of the entrepreneur CEO, successfully achieved those three stretch targets (low cost, high quality and prompt delivery), as shown by the superior financial performance. That performance was remarkable, with a significant growth in revenue from US$7.77 million in 2002 to US$144.34 million in 2006 (see Figure 2). Financial profit increased from −US$0.47 million in 2002 to US$3.30 million in 2006 (see Figure 3).

In sum, we investigated the organizational changes of the focal firm in a 12-year time period from 1995 to 2006, using CLDs and qualitative analysis. Based on that analysis, we can understand how managerial mental models change over time and how these changes affect strategic decision-making and organizational performance. Recall that, in Stage 1, the entrepreneur CEO had a firm and simple belief in product quality as the firm’s critical success factor, and in fact such insistence brought the company great success in the early stage. However, due to the firm’s continuing growth, coupled with radical changes in the business environment, senior management and the CEO perceived challenges, and gradually changed their mental models from a ‘quality-only’ one in Stage 1 to considering the three key successful factors (cost, quality and efficiency) together in Stages 2 and 3. According to the observed change in mental models, the firm made different strategic decisions in order to cope with the environmental challenges. Thus, organizational performance varied significantly at each stage. Overall, the firm achieved significant improvement in performance in a 12-year time period, as shown by an 1800% increase in profits from 1995 to 2006.

5. Discussion

To explain why there were significant performance changes within the firm under the same CEO and senior management team, this study has used a CLD approach to demonstrate how managerial mental models affect strategic responses and firm performance over time. This article makes important contributions to the mental model literature. We begin by summarizing the findings and implications for the mental model literature. Then we examine the implications of the findings for practice, study limitations and our conclusion.

Our study adds knowledge to the literature on mental models in three ways. First, while prior studies have largely examined managerial mental models from a static perspective (Mohammed et al., 2010; Roxas and Coetzer, 2012; Sterman et al., 2007), we unpack the dynamics of managerial mental models and trace the creation and evolution of mental models over a 12-year period of time. We reveal three distinct stages of change in managerial mental models with the case company. By doing so, we respond to recent calls for research on the longitudinal perspective of mental models and for examination of the formation and evolution of mental models over time (McComb, 2007; Mohammed et al., 2010). Our empirical results enable us to draw theoretically significant insights about the drivers of the changes in managerial mental models and the consequent effect on strategic choice and performance over time.

Second, our analysis implies that managerial mental models change along with entrepreneur’s beliefs and environmental competitiveness, and consequently these impact strategic decisions and organizational performance. The findings provide novel insights for the literature (Csaszar and Levinthal, 2015; Gary and Wood, 2011, 2016; Martignoni et al., 2016) by showing how individual factors and organizational factors simultaneously drive the changes and updates of managerial mental models and their dynamic effects on strategic choice and organizational performance. In particular, our study extends the recent research of Martignoni et al. (2016) on mental models. While their study suggests mental models and their interaction with environments as intermediate mechanisms between strategic choice and performance, our research further demonstrates that managerial mental models are changed based on the managerial perception of the environment dynamics and consequently affect performance.

Third, we track the changes in managerial mental models of the same senior management team over time. While most prior studies focus on the comparisons and changes in mental models from different management teams (Gavetti, 2005; Kaplan and Tripsas, 2008; Reger and Huff, 1993), we extend this line of research by examining the mental model dynamics in the same management team. We therefore also contribute to the research on cognitive change/inertia (Barr et al., 1992; Tripsas and Gavetti, 2000) by revealing the drivers of mental model dynamics of the same management team. While prior studies suggest that top management teams may have cognitive inertia and consequently fail to update their mental models in response to the business environment (e.g. Barr et al., 1992), the consistency (i.e. limited changes) of top management team members over time may help to explain why our investigated team successfully updated their managerial mental models. We believe that such consistency may reduce the development time of a shared mental model and increase work commitment. Nevertheless, we also acknowledge our empirical data do not allow us to explicitly suggest the difference between the case company and other similar companies regarding the changes in the managerial mental model. We thus recommend future research to further uncover other characteristics of top management teams that may affect the updating of their mental models.

The results of the analysis also provide important insights for practitioners. The primary problem with a managerial mental model is not whether it is right or wrong, but that it may exist below the level of awareness, deeply embedded in a firm’s management traditions and routines. From a long-term perspective, inappropriate mental models can eventually cause significant losses or result in great failure. However, the nature of the mental model imposes both theoretical and empirical challenges to the building of in-depth understanding of managerial mental models. It is unrealistic for managers to examine or reflect those shared beliefs, cognition and assumptions that have significant impact on the quality of firm decisions they make (Walsh, 1995). Thus our study provides practitioners with a systematic way to observe, unpack and examine the impact of different mental models of the firm.

Generalizability is an important element of all research and is a common issue with case studies. While the causal loop diagram based on the case company allows us to build in-depth understanding of the changes in mental models over time and the interactions of decision-making processes, our approach limits the generalizability of our findings (Sterman, 2000). The utility of our approach lies in its ability to build theoretical foundations to support larger scale data collection methods. Future research is recommended to mathematically formalize the relationships specified in the causal loop diagram (Repenning and Sterman, 2002; Sterman et al., 1997) and empirically test our causal relationships in a quasi- or laboratory-experimental study (Gary et al., 2017; Romanelli and Tushman, 1986; Yang et al., 2017).

While our work identifies the evolution of mental models, as represented in the causal loop diagrams, as a salient factor affecting the changes of organizational performance, we acknowledge that alternative explanations for the case company’s performance remain. For example, the global economy and the technology industry boom may both impose significant effects on the focal company’s performance. While we consider those alternative factors, we encourage future research to more quantitatively control for their potential effects.

6. Conclusion

Managers often rely on mental models to guide their decision-making. While prior studies have suggested the important role of mental models in firm strategies and performance, the understanding of mental models remains limited because of their dynamic nature (Walsh, 1995). We adopt a systematic approach to examining the impact of the managerial mental models over time. Our study implies that managerial mental models change with an entrepreneur’s beliefs and managers’ perceived environmental competitiveness and, consequently, these account for the changes in firm performance over time. Our study provides important implications for practitioners: decision-makers should observe and investigate their mental models before making strategic decisions.

Footnotes

Final transcript accepted 12 December 2018 by John Roberts (AE Strategy).

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The second-named author’s contribution to this work was sponsored by Peak Discipline Construction Project of Education at East China Normal University and Fundamental Research Funds for the Central Universities (2018ECNU-HWFW017).

Notes

ORCID iD

Miles M Yang

References

Bagdasarov

Johnson

MacDougall

et al . (2016) Mental models and ethical decision making: The mediating role of sensemaking. Journal of Business Ethics 138: 133–144.

Bandura

(1988) Organisational applications of social cognitive theory. Australian Journal of Management 13: 275–302.

Barr

Stimpert

Huff

(1992) Cognitive change, strategic action, and organizational renewal. Strategic Management Journal 13: 15–36.

Benner

Tripsas

(2012) The influence of prior industry affiliation on framing in nascent industries: The evolution of digital cameras. Strategic Management Journal 33: 277–302.

Boulding

Moore

Staelin

et al . (1994) Understanding managers’ strategic decision-making process. Marketing Letters 5: 413–426.

Cho

Hambrick

(2006) Attention as the mediator between top management team characteristics and strategic change: The case of airline deregulation. Organization Science 17: 453–469.

Chow-Chua

Goh

Boon Wan

(2003) Does ISO 9000 certification improve business performance? International Journal of Quality & Reliability Management 20: 936–953.

Csaszar

Levinthal

(2015) Mental representation and the discovery of new strategies. Strategic Management Journal 37: 2031–2049.

Davis

Eisenhardt

Bingham

(2007) Developing theory through simulation methods. Academy of Management Review 32: 480–499.

10.

Denrell

(2004) Random walks and sustained competitive advantage. Management Science 50: 922–934.

11.

Fassin

Werner

Van Rossem

et al . (2015) CSR and related terms in SME owner–managers’ mental models in six European countries: National context matters. Journal of Business Ethics 128: 433–456.

12.

Gary

Wood

(2011) Mental models, decision rules, and performance heterogeneity. Strategic Management Journal 32: 569–594.

13.

Gary

Wood

(2016) Unpacking mental models through laboratory experiments. System Dynamics Review 32: 99–127.

14.

Gary

Wood

Pillinger

(2012) Enhancing mental models, analogical transfer, and performance in strategic decision making. Strategic Management Journal 33: 1229–1246.

15.

Gary

Yang

Yetton

et al . (2017) Stretch goals and the distribution of organizational performance. Organization Science 28: 395–410.

16.

Gavetti

(2005) Cognition and hierarchy: Rethinking the microfoundations of capabilities’ development. Organization Science 16: 599–617.

17.

Gavetti

Levinthal

(2000) Looking forward and looking backward: Cognitive and experiential search. Administrative Science Quarterly 45: 113–137.

18.

Ginsberg

(1988) Measuring and modelling changes in strategy: Theoretical foundations and empirical directions. Strategic Management Journal 9: 559–575.

19.

Hagel

Rayport

(1997) The coming battle for customer information. Harvard Business Review 75: 64–77.

20.

Hall

(1984) The natural logic of management policy making: Its implications for the survival of an organization. Management Science 30: 905–927.

21.

Hodgkinson

Maule

Bown

(2004) Causal cognitive mapping in the organizational strategy field: A comparison of alternative elicitation procedures. Organizational Research Methods 7: 3–26.

22.

Kaplan

Tripsas

(2008) Thinking about technology: Applying a cognitive lens to technical change. Research Policy 37: 790–805.

23.

Karakaya

Yannopoulos

(2010) Defensive strategy framework in global markets. European Journal of Marketing 44: 1077–1100.

24.

Knight

Pearce

Smith

et al . (1999) Top management team diversity, group process, and strategic consensus. Strategic Management Journal 20: 445–465.

25.

Kroes

Manikas

(2014) Cash flow management and manufacturing firm financial performance: A longitudinal perspective. International Journal of Production Economics 148: 37–50.

26.

Lee

Faff

Langfield-Smith

(2009) Revisiting the vexing question: Does superior corporate social performance lead to improved financial performance? Australian Journal of Management 34: 21–49.

27.

Lincoln

Guba

(1985) Naturalistic Inquiry. Beverly Hills, CA: SAGE.

28.

McComb

(2007) Mental model convergence: The shift from being an individual to being a team member. In: Dansereau

Yammarino

(eds) Multi-Level Issues in Organizations and Time. Bingley: Emerald Group, pp. 95–147.

29.

Martignoni

Menon

Siggelkow

(2016) Consequences of misspecified mental models: Contrasting effects and the role of cognitive fit. Strategic Management Journal 37: 2545–2568.

30.

Mohammed

Ferzandi

Hamilton

(2010) Metaphor no more: A 15-year review of the team mental model construct. Journal of Management 36: 876–910.

31.

Morecroft

JDW

Sterman

(1994) Modeling for Learning Organizations. Portland, OR: Productivity Press.

32.

Moss

Stine

(1993) Cash conversion cycle and firm size: A study of retail firms. Managerial Finance 19: 25–34.

33.

Oliva

Sterman

(2001) Cutting corners and working overtime: Quality erosion in the service industry. Management Science 47: 894–914.

34.

Patton

(1980) Qualitative Evaluation Methods. Beverly Hills, CA: SAGE.

35.

Perlow

Okhuysen

Repenning

(2002) The speed trap: Exploring the relationship between decision making and temporal context. Academy of Management Journal 45: 931–955.

36.

Porac

Thomas

Baden-Fuller

(1989) Competitive groups as cognitive communities: The case of Scottish knitwear manufacturers. Journal of Management Studies 26: 397–416.

37.

Prahalad

Bettis

(1986) The dominant logic: A new linkage between diversity and performance. Strategic Management Journal 7: 485–501.

38.

Reger

Huff

(1993) Strategic groups: A cognitive perspective. Strategic Management Journal 14: 103–123.

39.

Reger

Palmer

(1996) Managerial categorization of competitors: Using old maps to navigate new environments. Organization Science 7: 22–39.

40.

Repenning

Sterman

(2002) Capability traps and self-confirming attribution errors in the dynamics of process improvement. Administrative Science Quarterly 47: 265–295.

41.

Romanelli

Tushman

(1986) Inertia, environments, and strategic choice: A quasi-experimental design for comparative-longitudinal research. Management Science 32: 608–621.

42.

Roxas

Coetzer

(2012) Institutional environment, managerial attitudes and environmental sustainability orientation of small firms. Journal of Business Ethics 111: 461–476.

43.

Rudolph

Repenning

(2002) Disaster dynamics: Understanding the role of quantity in organizational collapse. Administrative Science Quarterly 47: 1–30.

44.

Shea

Cooper

Cieri

et al . (2012) Evaluation of a perceived organisational performance scale using Rasch model analysis. Australian Journal of Management 37: 507–522.

45.

Simon

(1955) A behavioral model of rational choice. The Quarterly Journal of Economics 69: 99–118.

46.

Simon

(1991) Bounded rationality and organizational learning. Organization Science 2: 125–134.

47.

Sterman

(2000) Business Dynamics: Systems Thinking and Modeling for a Complex World. Boston, MA: McGraw-Hill.

48.

Sterman

Henderson

Beinhocker

et al . (2007) Getting big too fast: Strategic dynamics with increasing returns and bounded rationality. Management Science 53: 683–696.

49.

Sterman

Repenning

Kofman

(1997) Unanticipated side effects of successful quality programs: Exploring a paradox of organizational improvement. Management Science 43: 503–521.

50.

Terziovski

Samson

Dow

(1997) The business value of quality management systems certification. Evidence from Australia and New Zealand. Journal of Operations Management 15: 1–18.

51.

Thomas

(1988) Does leadership make a difference to organizational performance? Administrative Science Quarterly 33: 388–400.

52.

Tripsas

Gavetti

(2000) Capabilities, cognition, and inertia: Evidence from digital imaging. Strategic Management Journal 21: 1147–1161.

53.

Vandenbosch

Higgins

(1996) Information acquisition and mental models: An investigation into the relationship between behaviour and learning. Information Systems Research 7: 198–214.

54.

Walsh

(1995) Managerial and organizational cognition: Notes from a trip down memory lane. Organization Science 6: 280–321.

55.

Welch

(2003) The relationship between ownership structure and performance in listed Australian companies. Australian Journal of Management 28: 287–305.

56.

Yang

Jiang

Gary

(2017) Challenging learning goals improve performance in dynamically complex microworld simulations. System Dynamics Review 32: 204–232.

57.

Yang

Young

et al . (2016) Using system dynamics to investigate how belief systems influence the process of organizational change. Systems Research and Behavioral Science 34: 94–108.