We surveyed managers and other professionals in 142 US hospitality firms to investigate the links among relative firm size, market strategies (e.g. broad cost leadership and global growth) and nonmarket strategies. It helps explain how business size influences these dynamics within the hospitality industry. Previous research has primarily focused on generic market strategies without adequately considering the nuanced impact of firm size on performance outcomes.
The data are analyzed using PLS-SEM and machine learning (ML) methods, specifically the glmnet algorithm for Lasso regression, to explore the relationships between firm size, market strategies and performance. ML’s ability to manage high-dimensional data and nonlinear relationships provides a nuanced analysis that surpasses traditional statistical methods, enhancing the accuracy and depth of our findings.
The results depict a negative link between broad cost leadership and firm performance and a positive link between a global growth emphasis and firm performance. Firm size did not influence either market strategy or performance, but nonmarket orientation fully mediated the relationship between firm size and performance.
We assessed nonfinancial and financial performance with self-typing scales; objective measures can help evaluate strategy-performance linkages through a different lens and potentially reduce the influence of common method variance. Also, our assessment was limited to a cross-section of US hospitality firms. Additional work in the hospitality industry is required to identify and corroborate nonmarket strategies at the firm, strategic group and industry levels.
The current study employs both PLS-SEM and machine learning. It reinforces configuration theory by combining market and nonmarket methods as business performance indicators in hospitality firms.
A wealth of scholarship has preliminarily found links between market strategies and firm performance (Baron, 1995; Mellahi, Frynas, Sun, & Siegel, 2016). Market strategies concentrate on rivals, corporate resources, and skills to improve performance (Al-Surmi, Cao, & Duan, 2020; Varadarajan, 2020). In contrast, nonmarket strategies (NMS) can also enhance firm performance outside the market context (Topcuoglu, Kim, Kim, & Kim, 2022). A nonmarket orientation includes political and social interactions between organizations and external actors mediated by governments, public institutions, and other stakeholders (Baron, 1995). Scholarship on the social and political dimensions of nonmarket strategy among hospitality firms has increased recently (Eid, Agag, & Shehawy, 2021; Uyar, Kilic, Köseoglu, Kuzey, & Karaman, 2020).
NMS differs from market strategies that depend on financial tactics, including price, quality of products, and branding, to improve business performance (Xiao, Chan, & Chen, 2023). As a result, several corporations participate in significant nonmarket activities, often categorized as social or political (Mellahi et al., 2016). Social NMS contains corporate social responsibility (CSR) and associated public behaviors (Kamasak, James, & Yavuz, 2019; Liedong, 2022), such as virtue-signaling by purchasing carbon credits. Political NMS often encompasses less transparent actions such as lobbying or hiring the friends and relatives of influential people (Parnell & Brady, 2019). Recent research underscores the importance of nonmarket activity as a consequential part of a firm’s strategy and a potential driver of performance (Parnell, 2018; Sun, Doh, Rajwani, & Siegel, 2021).
Scholarship investigating the strategy-performance relationship in the hospitality industry has evolved considerably in recent decades (Köseoglu, Sehitoglu, Ross, & Parnell, 2016). Some of this work evaluates corporate social responsibility (CSR) as a performance driver (Köseoglu et al., 2020, 2022; Serra-Cantallops, Peña-Miranda, Ramón-Cardona, & Martorell-Cunill, 2018). Researchers have also discovered a correlation between NMS and performance, although competing explanations exist (Parnell, 2018), especially among hospitality firms (Köseoglu et al., 2020).
However, there are several gaps in the literature. Although the corpus on performance indicators is extensive (e.g. Harris & Mongiello, 2001; Ribeiro, Vasconcelos, & Rocha, 2019), relative business size is not commonly considered a driver, especially in the hospitality industry. This deficit has been underscored by Bititci, Garengo, Dörfler, and Nudurupati (2012), who performed a meta-review and discovered that the traditional independent variables in strategic management do not explain enough about performance in small firms. While existing literature has extensively explored market strategies and firm performance, this study identifies a gap in understanding how business size influences these dynamics within the hospitality industry. Previous research has primarily focused on generic market strategies without adequately considering the nuanced impact of firm size on performance outcomes. Additional work is essential, given that theoretical frameworks such as the resource-based view and industrial organization theory suggest firm size plays a pivotal role in shaping competitive advantages through economies of scale, market power, diversification, and risk management capabilities. For instance, larger firms can leverage their size to negotiate better terms with suppliers, access broader markets, and diversify their investment portfolio to manage risks more effectively. These theoretical linkages suggest a complex relationship between firm size and performance indicators, which remains underexplored in the context of the hospitality industry. Hence, a firm’s relative size is essential because small or micro hospitality firms may have distinct features that impact their market strategies and performance levels (López-Fernández, Serrano-Bedia, & Gómez-López, 2011). Company size influences entrepreneurial behavior in tourism (Kallmuenzer & Peters, 2018). By addressing this gap, our study aims to elucidate the nuanced mechanisms through which business size impacts firm performance, thereby contributing to a more comprehensive understanding of strategic management in this sector.
We consider whether market strategies (operationalized as broad cost leadership and global growth strategic orientation) and NMS mediate the relationship between relative firm size and performance. We maximize insight by employing both PLS-SEM to examine linear relationships (which might be limiting) between variables and machine learning methods to assess both linear and non-linear relationships. We used machine learning (ML) methods to elucidate the complex relationship between firm size, market strategies, and performance indicators in the industry. ML, a subset of artificial intelligence, analyzes data to learn patterns and make predictions. Unlike traditional statistical methods that may struggle with high-dimensional data or nonlinear relationships, ML methods can effortlessly handle these complexities. This capability is particularly advantageous in our study, where the intricate dynamics of the hospitality industry demand a nuanced analysis that goes beyond the capabilities of conventional approaches. Among various ML methods, we chose the glmnet algorithm to conduct Lasso regression due to its robust performance in handling specific data characteristics (e.g. multicollinearity, nonlinearity) and its proven track record in similar studies. These algorithms allow for a more accurate and comprehensive analysis of the data, providing insights that traditional methods may overlook.
Theoretical background and hypothesis development
Our approach is grounded in three theoretical perspectives. First, configuration theory (Kreiser, Kuratko, Covin, Ireland, & Hornsby, 2021) investigates how different combinations of strategic and structural variables influence performance (Sheehan & Foss, 2007). Assessing configurations focuses on patterns of distinct strategies or characteristics that often co-occur and explicates the reality of strategic equifinality (Donaldson, 2001). Configuration theory posits that successful firms achieve superior performance by creating coherent configurations that suit their specific environments and contexts. It also highlights the need for adaptability; as firms encounter environmental changes, they must reconfigure their strategies to maintain alignment. Thinking about configurations helps strategies frame strategic decisions and make informed decisions that enhance their competitive advantage and operational efficiency.
Firms should align their nonmarket strategies with other contextual variables. For example, large firms—especially those operating in highly regulated or politically sensitive industries—can leverage NMS more because their performance is directly influenced by policy changes, government decisions, and public sentiment. However, NMS offers diminishing returns for firms in less regulated environments and smaller ones with limited access to resources (Liedong, 2022; Parnell, Troilo, & Dobbelstein, 2024). We focus on strategic intent and seek to explain how emphasizing market and nonmarket strategies drive firm performance (Fiss, 2007).
Second, the notion of NMS is grounded in public choice theory, which explains why organizations seek favorable arrangements with government stakeholders (Buchanan & Tullock, 1962; Gilli, Li, & Qian, 2018). Public choice theory views firms as rational actors that lobby, contribute to political campaigns, or seek to influence regulations when the expected benefits outweigh the costs. Hence, larger firms are more likely to employe NMS because they have more at stake and can better absorb the costs of political engagement (Congleton, 2018). We test the public choice perspective by investigating a link between NMS and firm performance among U.S. hospitality firms.
Third, NMS is grounded in social exchange theory, which can explain organizational decisions as cost-benefit assessments (Homans, 1958; Priporas, Stylos, Rahimi, & Vedanthachari, 2017). From this perspective, firms engage in NMS because they believe the benefits (e.g. an attractive regulatory regime, positive responses from customers) exceed the costs (e.g. lobbying, negative responses from customers).
Firm size
Firm size is “a proxy for total and slack resources that represents the organization’s economies of scales” (Lee & Xia, 2006, p. 975). It can influence its emphasis on strategy and stakeholder orientation, but it is commonly treated as a control variable (Parnell & Brady, 2019). Although control variables are presumed to influence the dependent variable, which is why they are “controlled for” in the analysis, they can be overlooked as an independent variable. Firm size is often more than tangentially associated with performance (Davis & Bendickson, 2021; Wong, Wong, & Boon-itt, 2020).
While larger firms have the resources and the ability to spend on R&D, such expenditures are more difficult for small and medium-sized hospitality firms (Wong et al., 2020). They also have economies of scale when purchasing mattresses, linens, and all other items in a hotel or resort (Youn, Hua, & Lee, 2015). Researchers frequently include organizational scale an essential indicator of corporations’ market strategy. Companies operating in a particular market will likely differ in size, resources, and ambitions. Some are enormous, have a lot of resources, and want to control the industry; others are small, have few resources, and merely want to thrive in a limited way, such as having hotels only in one geographical region. Because of these differences, enterprises in the target market pursue distinct tactics (Desai, 2013).
Firm size is often associated with firm growth for several reasons. Larger businesses are more likely to have more resources, improving their capacity to acquire and analyze information and giving the firm a competitive edge (Davis & Bendickson, 2021; Wong et al., 2020). Also, scale economies in internationally focused investments has been linked to business size (Li, Zhang, & Shi, 2020).
As a result, business size often drives enterprises to become more internationally focused. However, decades of research on the influence of business size have produced conflicting results (Li et al., 2020). Some scholars have demonstrated a negative business size-growth tendency (e.g. Coad & Hölzl, 2009). Small organizations often attempt to expand more rapidly than large firms because growth engenders scale economies (Park & Jang, 2010).
Broad cost leadership
Porter’s (1985) generic strategy typology suggests that a business can obtain competitive advantage and super performance through either cost leadership (e.g. Red Roof Inn) or differentiation (e.g. Distinctive Hospitality Group). Cost leadership can enhance performance by reducing costs relative to rivals, permitting a firm to lower prices while maintaining reasonable margins. Substantial research has supported a positive link between cost leadership and firm performance (Brenes, Montoya, & Ciravegna, 2014; Lee, Hoehn-Weiss, & Karim, 2021). Cost leadership is often buttressed by economies of scale and usually pursued by large firms (Banker, Mashruwala, & Tripathy, 2014; Lee et al., 2021). These arguments imply that firm performance depends on cost leadership, which is impacted directly by a firm’s size relative to its competitors.
Early studies encouraged applying Porter’s notion of strategic groups (Bradburd & Ross, 1989; Gibcus & Kemp, 2003). Whereas the techniques employed by small and medium-sized enterprises (SMEs) differ, new and small businesses often find their greatest success in pursuing cost leadership (Parnell, Lester, Zhang, & Köseoglu, 2012). Due to limited resources, numerous small firms focus on niche markets, whether geographical, product-based, or service-based, while stressing cost leadership (Thomas, 2000). Small cost leaders that enter significant markets face ongoing pressure from huge, established brands and businesses, placing them at a considerable disadvantage.
Firm size and broad cost leadership
Though various approaches provide diverse perspectives on the interaction between business size and market strategy, research in the hospitality industry is limited. Scholars have yet to identify a consistent link between firm size and market tactics. Although most small companies do not have the means and know-how to strengthen market achievement, their size gives them greater flexibility and independence from organizational bureaucracy (Leal-Rodríguez, Eldridge, Roldán, Leal-Millán, & Ortega-Gutiérrez, 2015), sometimes improving performance. Hence, we suggest the following:
Relative firm size will be positively associated with its emphasis on broad cost leadership.
A firm’s emphasis on broad cost leadership will be positively associated with its performance.
Firm size and growth orientation
The international expansion of hospitality enterprises, such as InterContinental Hotels, America’s Best Value, and the Ritz-Carlton, has been a consistent trend. Even during recessions, some firms seek to diversify geographic exposure (García-Almeida & Yu, 2015). Goals relating to profitability and growth are critical in the hotel business (O’neill & Mattila, 2006). Johnson and Vanetti (2007) used the international product life cycle model to demonstrate that domestic market saturation prompted the global expansion of hospitality companies.
Because hotels and restaurants are in areas where consumption occurs, the development patterns of companies in the hospitality industry vary markedly from those of other businesses. Therefore, a hotel can raise its capacity somewhat, but profit may be jeopardized when a certain number of rooms or food and beverage consumption points are reached (Brown & Dev, 1999; Burgess & Bryant, 2001). This capacity limitation opens the door to tackling the primary growth strategy in this industry by integrating additional units through geographic diversity (García-Almeida & Yu, 2015).
This study postulates a positive association between company size and global expansion orientation for the hotel industry to elucidate specific weak and inconsistent results. In other words, large businesses often develop faster than small firms globally, or small firms grow more slowly than large firms. We also hypothesize that a global, growth strategic orientation influences a firm’s performance. Formally stated:
Relative firm size will be positively associated with a global, growth strategic orientation.
A firm’s global, growth strategic orientation will be positively associated with its performance.
Firm size and nonmarket strategy
A firm’s size, structure, and circumstances can drive its nonmarket orientation (Marzouk, 2017). Indeed, firms pursue social and political nonmarket strategies to improve performance (Liedong, Aghanya, & Rajwani, 2020; Wrona & Sinzig, 2018). The link between NMS and performance is intuitive and multifaceted (Parnell, 2015). A nonmarket orientation seeks to build relationships with stakeholders; firms would not pursue NMS if they did not expect it to enhance performance. Most published work suggests a link between the political and social dimensions of NMS (e.g. Hadani & Coombes, 2015; Marquis & Raynard, 2015), but there are some exceptions (e.g. Cho, Laine, Roberts, & Rodrigue, 2018; Liedong, Rajwani, & Mellahi, 2017).
NMS can drive performance in many ways (Hadani & Schuler, 2013; Mellahi et al., 2016). Social NMS can increase overall firm performance by helping the organization achieve broader social objectives (Bosse, Phillips, & Harrison, 2009; Eid et al., 2021; Harrison & Wicks, 2013; Jadnanansing, DiPietro, & De Droog, 2024). Moreover, widely-accepted stakeholder theory highlights the strategic importance of entities beyond shareholders and customers (Hillman & Keim, 2001).
Scholars have identified positive, direct links among stakeholder management (Bosse et al., 2009; Choi & Wang, 2009), social interaction (Alhouti, Johnson, & Holloway, 2016; Mbalyohere & Lawton, 2018), broad nonmarket activity (Bonardi, Holburn, & Vanden Bergh, 2006; Parnell, 2015), and performance. In their review of scholarship on the NMS-performance link, Mellahi et al. (2016) found that 102 out of 163 studies evaluating a form of NMS and performance identified a positive association.
Nonmarket tactics and exchanges with various parts of the institutional environment are influenced by firm characteristics such as size and “the ‘rules of the game’ controlling economic interaction” (Dorobantu et al., 2017, p. 114). Sanusi and Connell (2018) found that a firm’s size, business climate, and government activity impact its NMS.
Large firms tend to benefit more by employing political and social NMS (Liedong et al., 2020; Liedong, Ghobadian, Rajwani, & O'Regan, 2015). Their negotiation with politicians and regulators demonstrates they already have access to the political system. Social NMS can facilitate political NMS by fostering goodwill and conferring legitimacy (Rodgers, Stokes, Tarba, & Khan, 2019). Hence, we proffer the following:
Relative firm size will be positively associated with its nonmarket orientation.
A firm’s nonmarket orientation will be positively associated with its performance.
Hypothesized mediating relationships
Though numerous approaches give diverse perspectives on the connection between company size and profitability, empirical evidence on the relationship between these two dimensions is often weak or conflicting (e.g. Mas-Ruiz & Ruiz-Moreno, 2011; Treen, Atanasova, Pitt, & Johnson, 2016; Zhang, Parnell, & Xiong, 2020). Pervan and Višić (2012) identified positive, negative, and nonsignificant impacts.
Firm size can influence the relationship between strategy and performance (Kallmuenzer & Peters, 2018). For example, small firms may not possess the know-how to enhance innovative performance. Still, their size provides greater flexibility and independence from institutional bureaucracy (Leal-Rodríguez et al., 2015), improving performance. On the other hand, larger enterprises, such as worldwide hotel chains, are typically capable of adopting accountability by establishing processes that enable possible transactions with local government and civil society (i.e. NMS) with well-defined goals, metrics, and processes (Youn et al., 2015). Hence, large hospitality firms may be better positioned to create more robust market and nonmarket strategies.
These contradicting results prompt academics to investigate a potential mechanism that might relate to the influence of firm size on industry company performance. In other words, a scholarly understanding of the mechanisms that support or impede firm success is far from comprehensive, with some critical elements lacking. Various structural and strategic factors could mediate the link between size and performance (Kharub, Mor, & Sharma, 2019; Whetten, 1989). Hence, we anticipate that market and nonmarket strategies will mediate the link between company size and firm performance among hospitality firms.
The relationship between relative firm size and firm performance will be mediated by an emphasis on broad cost leadership.
The relationship between relative firm size and firm performance will be mediated by a firm’s emphasis on global growth.
The relationship between relative firm size and firm performance will be mediated by a firm’s nonmarket orientation.
Figure 1 illustrates the hypothesized links among relative firm size, market strategy, nonmarket strategy, and firm performance. Firm age is a control variable.
Methods
Sample and questionnaire
We used the (Eyal, David, Andrew, Zak, & Ekaterina, 2021; Tang, Birrell, & Lerner, 2022) platform to survey 142 managers and other professionals in the U.S. hospitality industry. Prolific’s targeting capabilities permit the detailed identification and recruitment of participants based on demographic and other specified qualifications. Prolific data have been used in a variety of scholarly research. Its validity as an academic platform has received support in the literature (Eyal et al., 2021; Tang et al., 2022).
In addition to standard demographics, the survey included questions about relative firm size, number of employees, broad cost leadership, global growth orientation, nonmarket emphasis, and firm performance. Managers throughout the organization were included in the analysis, as they have recently played a greater role in strategy formulation and execution (Balogun & Johnson, 2004; Raes, Heijltjes, Glunk, & Roe, 2011). We removed responses that were completed too quickly, included evidence of straightlining, or contained more than 10% missing data. Table 1 provides a summary of the respondents and their organizations.
The sample
| Variable | n = 142 | % |
|---|---|---|
| Position | ||
| Professionals | 20 | 14.1 |
| Supervisory manager | 48 | 33.8 |
| Middle manager | 52 | 36.6 |
| Top manager | 22 | 15.5 |
| Gender | ||
| Male | 78 | 54.9 |
| Female | 64 | 45.1 |
| Firm size | ||
| Micro (1–10 employees) | 19 | 13.4 |
| Small (11–50 employees) | 32 | 22.5 |
| Medium (51–250 employees) | 24 | 16.9 |
| Large (251+ employees) | 67 | 47.2 |
Source(s): Authors’ own work
We measured relative firm size by asking respondents to compare the size of their firms to those of their competitors (i.e. much smaller, smaller, about the same size, larger, or much larger). This decision is grounded in the notion that relative firm size offers critical insights into a firm’s competitive standing and strategy formulation processes. Unlike absolute size, which quantifies a firm’s scale in isolation, relative size positions a firm within its competitive landscape, providing a lens through which strategic decisions and performance outcomes can be better understood.
The categorization of relative firm size facilitates an investigation into strategic orientations from a competitive perspective. This approach acknowledges that managers and industry professionals often perceive and react to their firm’s size in a relational context, making strategic choices based on competitive positioning rather than absolute metrics. The use of categorical variables in this context not only aligns with our research objectives but also enhances the practical relevance of our findings by mirroring the decision-making processes within the industry. Theoretical support for this approach is found within competitive dynamics theory, which emphasizes the importance of relative positioning in competitive strategy. Empirically, studies have demonstrated the utility of relative size measures in capturing the strategic dynamics and performance implications within competitive industries, including hospitality. These precedents underscore our rationale for employing a categorical measure of relative firm size, facilitating a nuanced exploration of how firms navigate their competitive environments.
We used 12 Likert items to assess the extent to which respondents’ organizations employ competitive (market) and nonmarket strategies and perform relative to their rivals (see Table 2). We measured market strategy with nine items based on previous research (see Kellermanns & Eddleston, 2006; Parnell et al., 2024; Zhang et al., 2020). The market strategy items coalesced around two themes, broad cost leadership and global growth orientation. Nonmarket strategy was measured with items that assess the social and political dimensions, as well as an overall evaluation. Our firm performance measure includes three items: financial, non-financial, and overall performance (see Parnell, 2018). Most published strategy-performance studies have focused on financial indicators or related outcomes such as risk reduction or competitive advantage (Brozovic, 2018). However, research on the balanced scorecard and stakeholder management concepts underscores the importance of non-financial measures such as employee satisfaction, customer satisfaction, and capability development (Jusoh & Parnell, 2008; Köseoglu, Parnell, & Topaloglu, 2013; Parnell, 2021).
Survey items – reflective measures
| Item | Loading | Content |
|---|---|---|
| Broad cost leadership (α = 0.643, composite reliability = 0.722, AVE = 0.564) | ||
| Broad | 0.544 | Focus on a broad group of customers (recoded) |
| Cost | 0.804 | Minimizing costs |
| Not differ | 0.865 | Producing unique goods and services (recoded) |
| Global growth orientation (α = 0.777, composite reliability = 0.854, AVE = 0.681) | ||
| Global | 0.885 | Pursuing opportunities outside of our home country |
| Growth | 0.868 | Growing the organization |
| Profit | 0.712 | Maximizing profits |
| Nonmarket emphasis (α = 0.630, composite reliability = 0.700, AVE = 0.571) | ||
| CSR | 0.869 | Promoting social responsibility |
| Government | 0.631 | Working closely with governments, politicians, and regulators |
| Overall | 0.749 | Seeking to improve organizational performance through involvement in social, community, political, and government activities |
| Firm performance (α = 0.834, composite reliability = 0.839, AVE = 0.753) | ||
| Financial | 0.824 | Financial performance |
| Non-finan. | 0.841 | Non-financial performance |
| Overall | 0.933 | Overall performance |
Source(s): Authors’ own work
Our methodological approach included both structural equation modeling and machine learning. We used partial least squares structural equation modeling with SmartPLS version 4 to test the hypotheses (Hair, Howard, & Nitzl, 2020; Sarstedt, Ringle, & Hair, 2021). SmartPLS has been used in many hospitality studies (Chong, Lee Peng, & I-Chi, 2024; Murillo-Ramos, Huertas-Valdivia, & García-Muiña, 2024; Shehawy, 2022; Smith, White-McNeil, & Ali, 2020).
Measurement
We followed established guidelines when evaluating the measurement and structural models (Hair, Sarstedt, Ringle, & Gudergan, 2024). The constructs were measured reflectively. Reliability and validity were assessed with the partial least squares (PLS) algorithm (see Table 2). Construct reliability was assessed with Cronbach’s alpha (Nunnally, 1978). Scores exceeded 0.600 in all instances and 0.700 with two exceptions. The two instances where alpha scores were between 0.600 and 0.700 were three-item scales that exceeded recommendations for composite reliability and average variance explained (AVE). Composite reliability exceeded 0.700 (Hair et al., 2024) and AVE scores exceeded 0.500 for all constructs (Ashill, Carruthers, & Krisjanous, 2005). Hence, the measures were deemed reliable overall.
The Fornell-Larcker matrices presented in Table 3 suggest discriminant validity in all constructs (Sleimi & Emeagwali, 2017) and are reinforced by the heterotrait-monotrait (HTMT) output presented in Table 4. Discriminant validity is established when HTMT values are below 0.85. Moreover, none of the confidence intervals include the corresponding threshold values (or a more conservative value of 0.85) (Franke & Sarstedt, 2019).
Fornell-Larcker matrix
| Broad cost leadership | Nonmarket emphasis | Firm performance | Global growth | |
|---|---|---|---|---|
| Broad cost leadership | 0.751 | |||
| Nonmarket emphasis | −0.291 | 0.756 | ||
| Global growth | −0.177 | 0.225 | 0.395 | 0.825 |
| Relative firm size | −0.099 | 0.242 | 0.166 | 0.098 |
Source(s): Authors’ own work
Heterotrait-monotrait (HTMT) ratio
| Broad cost leadership | Nonmarket emphasis | Firm age | Firm performance | Relative firm size | |
|---|---|---|---|---|---|
| Broad cost leadership | 0.401 | ||||
| Nonmarket emphasis | 0.032 | 0.329 | |||
| Firm age | 0.664 | 0.607 | 0.066 | ||
| Firm performance | 0.271 | 0.332 | 0.053 | 0.465 | |
| Relative firm size | 0.103 | 0.296 | 0.094 | 0.182 | 0.118 |
Source(s): Authors’ own work
Common method bias (CMB) can distort the relationships between variables, particularly in structural models. Minimizing CMB helps ensure that the findings accurately reflect true relationships rather than artifacts of the measurement (Jordan & Troth, 2020; Podsakoff, Podsakoff, Williams, Huang, & Yang, 2024). We employed two tests to investigate possible CMB. First, the results from Harman’s single-factor test suggest that one factor accounts for 31.001% of the variance. Second, factor-level VIF scores were less than 3.3 in all instances. These results suggest that the model is relatively free from common method bias (Kock, 2015).
We examined item-level variance inflation factor (VIF) scores to identify possible collinearity influences. The VIF scores were below 3.0 for all items, with one exception. The overall firm performance item was 3.180, very close to 3.0. Overall, these results suggest collinearity is not a significant concern.
We designed the study model, including hypothesized relationships, to reveal potential factors affecting firm performance of hospitality organizations. We conducted advanced regression analysis with machine learning algorithms to support our model.
Our work intends to build algorithms that categorize determinants of business performance based on firm size, market, and nonmarket strategies using machine learning techniques. This research identified important research voids in predicting hospitality firms’ performance. Our study is novel because it builds on PLS-SEM results by employing machine learning to increase the detection of firm performance in internal and external contexts, two areas where firms have a unique potential to find elements impacting their performance. Because few or no machine learning studies have employed a second, independent sample to validate models, this research strategy represents a unique potential to advance the science of machine learning research. Our machine learning models are expected to outperform linear regression methods in categorizing company performance indicators in hotel contexts. Building on scholarship applying machine learning to organizations (e.g. Brei, 2020; Wang, Jia, Chen, & Xu, 2022), we sought to discover the most important drivers of firm performance and better understand the relationship among variables in the model.
To identify important predictors, we used R programming and simple forward selection methods to perform multiple regression and best subset regression. The regularized regression model, which included LASSO regression, was the second model into which we integrated predictors. Lasso regression is beneficial for decreasing the dimensionality of variables in a dataset and selecting appropriate forecasting factors (Wang, Li, & Jiang, 2007). To estimate, the lasso model minimizes the following function based on the standard regression model βj:
where n represents the number of episodes and i denotes the index of each observation in the regression. As illustrated in the equation, the minimizing tool incorporates the regularization term to preserve modest regression coefficients βj. The penalty level is indicated by the regularization term λ. Lasso regression produces an ordinary least squares (OLS) estimate when λ = 0. The penalty term, mirrored in the equation’s parameter, causes some coefficients to be approximated as zero. As a result, lasso regression is a popular method for choosing variables. In a lasso regression, the value of λ is crucial (Tian, Yang, Mao, & Tang, 2021). We used two methods to choose a suitable value: best lambda and bestclose lambda. Following that, we used relaxed lasso regression to complete the model.
In employing the LASSO regression method, we aimed for a rigorous yet exploratory approach to identify key predictors of firm performance within the hospitality sector. Unlike traditional OLS regression, which treats all variables equally without regard to multicollinearity or overfitting, LASSO introduces a penalty term that constrains the size of coefficients, effectively selecting a subset of predictors by reducing others to zero. This method mitigates the risk of overfitting—a common concern in models with numerous predictors—and provides a clear indication of which variables most strongly influence firm performance, thereby enhancing the model’s predictive accuracy.
Each of the 13 indicators was carefully chosen based on theoretical considerations and empirical evidence suggesting their relevance to firm performance. Including firm age is supported by literature indicating that the maturation of firms can influence their strategic flexibility, market adaptation, and operational efficiencies—factors crucial in the competitive hospitality industry.
In comparing LASSO with SEM, it’s important to clarify that while SEM excels at modeling complex relationships between latent constructs, LASSO is an effective preliminary step for variable selection. This complementary use of LASSO ensures that SEM analyses focus on the most pertinent variables, thus avoiding the “kitchen-sink model” pitfall and enhancing our findings’ overall robustness and interpretability.
To ensure the robustness of our feature selection process, we employed LASSO regression, renowned for its capacity to reduce overfitting by penalizing the absolute size of the regression coefficients. This approach not only aids in selecting features that contribute most significantly to our model but also helps in managing multicollinearity among predictors. Given the constraints posed by our sample size, traditional data splitting into training and test sets was deemed impractical for our ML approach. Splitting a small dataset can produce non-representative training and test samples, skewing model performance metrics. To circumvent this issue and still rigorously assess our model, we employed cross-validation, allowing us to maximize our available data for training while still providing a reliable estimation of the model’s predictive performance. This approach is particularly suited for studies like ours, where sample size limitations preclude using a separate test set but where model validation remains paramount.
In summary, our study employs a dual-method approach to analyze the strategic orientation and performance of hospitality firms. We utilize Partial Least Squares Structural Equation Modeling (PLS-SEM) to assess linear relationships and machine learning techniques to uncover nonlinear interactions. This comprehensive approach allows us to capture the complexity of the relationships between firm size, market strategies, nonmarket strategies, and firm performance.
We employed three independent variables. (1) Firm size was measured by respondents’ assessment of their firm size relative to competitors. (2) Market strategies include broad cost leadership (focusing on minimizing costs and targeting a broad customer base) and global growth orientation (pursuing international opportunities and maximizing profits). (3) NMS includes political and social interactions, CSR, and government engagement. Our dependent variable, firm performance, was evaluated using financial performance (e.g. revenue, profit margins), non-financial performance (e.g. customer satisfaction, employee satisfaction), and overall performance. We included firm age as a control variable. It is measured in years and is considered in our analysis to control for its impact on strategic orientation and performance.
Findings
Hypothesis testing
The bootstrapping algorithm in SmartPLS was utilized to test each hypothesis. Effect size was assessed with f2 values and interpreted following Cohen’s (1988) benchmarks of 0.02 (small), 0.15 (moderate), and 0.35 (large). The path model presented in Figure 2 includes path coefficients, p-values, and effect sizes. The circles for each dependent variable in the model contain R2 values. The results of the hypothesis tests are provided in Table 5.
Tests of hypotheses
| Hypothesis | Orig. sample | Sample mean | Std. dev | t-stat | p-value | Support | f2 value |
|---|---|---|---|---|---|---|---|
| H1a: Size → Cost | −0.099 | −0.101 | 0.088 | 1.123 | 0.262 | No | 0.010 |
| H1b: Cost → Perform | −0.423 | −0.423 | 0.062 | 6.852 | 0.000 | Yes | 0.317 |
| H2a: Size → Global growth | 0.098 | 0.098 | 0.085 | 1.158 | 0.247 | No | 0.010 |
| H2b: Global growth → Perf | 0.251 | 0.248 | 0.068 | 3.667 | 0.000 | Yes | 0.112 |
| H3a: Size → Nonmarket | 0.243 | 0.244 | 0.079 | 3.073 | 0.002 | Yes | 0.063 |
| H3b: Nonmarket → Perf | 0.294 | 0.300 | 0.067 | 4.357 | 0.000 | Yes | 0.132 |
| H4a: Size → Cost → Perf | 0.042 | 0.043 | 0.038 | 1.099 | 0.272 | No | n/a |
| H4b: Size → Glob growth → Perf | 0.025 | 0.025 | 0.023 | 1.075 | 0.282 | No | n/a |
| H4c: Size → Nonmarket → Perf | 0.071 | 0.074 | 0.030 | 2.347 | 0.019 | Yes | n/a |
Source(s): Authors’ own work
The first hypothesis was not supported. The link between relative firm size and broad cost leadership (H1a) was not significant. The link between broad cost leadership and firm performance (H1b) was significant but negative. The effect size (0.317) was moderate and approaching large.
The second hypothesis was partially supported. The link between relative firm size and the emphasis on global growth (H2a) was not significant. However, the link between global growth and firm performance (H2b) was positive and significant, with a moderate effect size (0.112).
The third hypothesis was supported. The link between relative firm size and nonmarket orientation (H3a) was positive and significant. The link between nonmarket orientation and firm performance (H3b) was also positive and significant, with a moderate effect size (0.132).
The fourth hypothesis was partially supported. The relationship between relative firm size and firm performance was not mediated by broad cost leadership (H4a) or global growth (H4b), but it was fully mediated by nonmarket orientation (H4c).
A model including all the hypothesized relationships was compared to a saturated model. No significant links were identified in the saturated model. The Bayesian information criteria (BIC) calculations for each dependent variable in the proposed model were below those in the saturated model, providing overall support for the proposed model.
Exploring the predictors
Table 6 demonstrates descriptive patterns of predictor items for firm performance. We removed all missing data from the sample, resulting in 111 participants remaining. We used factor scores of two variables (FPNM and FSNM) by subjecting their items to factor analysis because inter-correlation coefficient values between these two variables are very high. As can be observed, the results confirm normal distribution of the data as skewness and kurtosis values ranged between + −3.
Descriptive patterns of predictors
| Items | Codes | Mean | S.D | Trimmed | Skewness | Kurtosis |
|---|---|---|---|---|---|---|
| Str.Cost | x1 | 3.66 | 1.13 | 3.8 | −1.03 | 0.3 |
| Str.Unique | x2 | 3.46 | 1.29 | 3.52 | −0.38 | −1.22 |
| Str.Focus | x3 | 2.81 | 1.31 | 2.77 | 0.27 | −1.26 |
| Str.Growth | x4 | 3.62 | 1.19 | 3.69 | −0.55 | −0.92 |
| Str.Global | x5 | 3.26 | 1.41 | 3.32 | −0.31 | −1.28 |
| Str.Profit | x6 | 4.04 | 1.04 | 4.22 | −1.22 | 1.01 |
| Str.Gov | x7 | 2.23 | 1.19 | 2.13 | 0.5 | −0.99 |
| Str.CSR | x8 | 2.95 | 1.26 | 2.93 | −0.09 | −1.16 |
| NMS overall | x9 | 3.11 | 1.25 | 3.13 | −0.26 | −1.16 |
| FPNMS | x10 | 0.08 | 1.01 | −0.04 | 0.74 | −0.56 |
| FSNMS | x11 | 0.05 | 1 | 0 | −0.14 | −0.88 |
| Firm age | x12 | 45.31 | 31.68 | 41.68 | 0.88 | 0.11 |
| OrgSizeComp | x13 | 2.96 | 1.2 | 2.94 | 0.05 | −0.83 |
| Performance | y | 3.54 | 1.06 | 3.58 | −0.55 | −0.65 |
Source(s): Authors’ own work
Figure 3 shows correlation coefficients between predictor and outcome variables. Correlations between x2 and y (r = 0.528, p < 0.001) and x11 and y (r = 0.254, p < 0.001) are positive and significant, while there is a negative and significant correlation between x1 and y (r = −0.413, p < 0.001). However, x3 (r = 0.132, p > 0.05) and x10 (r = 0.113, p > 0.05) do not have a significant relationship with the outcome variable. More importantly, the diagonal matrix (i.e. snapshots) has shown that the correlations are primarily non-linear.
First, we applied multiple regression. The results reveal four predictors having a significant effect on outcomes variable (i.e. firm performance) (see Table 7). Among them, x1 has a significant negative impact on firm performance (t = −2.383, p < 0.05), whereas x2 (t = 3.88, p < 0.001), x7 (t = 2.176, p < 0.05), and x8 (t = 2.53, p < 0.05) have a significant positive impact on the outcome variable.
Multiple regression analysis
| Estimate | S.E | t-statistic | p-value | |
|---|---|---|---|---|
| Intercept | 1.63 | 0.73 | 2.21 | 0.028 |
| X1 | −0.2 | 0.08 | −2.38 | 0.019 |
| X2 | 0.28 | 0.07 | 3.88 | 0 |
| X3 | −0.05 | 0.06 | −0.79 | 0.43 |
| X4 | 0.16 | 0.1 | 1.64 | 0.103 |
| X5 | −0.04 | 0.08 | −0.56 | 0.576 |
| X6 | 0.02 | 0.1 | 0.18 | 0.85 |
| X7 | 0.21 | 0.09 | 2.17 | 0.031 |
| X8 | 0.2 | 0.08 | 2.53 | 0.013 |
| X9 | 0.04 | 0.09 | 0.45 | 0.65 |
| X10 | −0.2 | 0.11 | −1.71 | 0.089 |
| X11 | −0.06 | 0.14 | −0.43 | 0.667 |
| X12 | −0.01 | 0.01 | −0.79 | 0.428 |
| X13 | 0.05 | 0.07 | 0.82 | 0.412 |
Note(s): S.E. denotes standard error
Source(s): Authors’ own work
Second, we utilized the basic variable selection method – forward step. To formulate the best model, we consider two selection criteria – BIC and Adjusted R2 Criterion (ADJR2). Figure 4 shows how many predictors provide the best model. While for BIC three predictors generate the best model for ADJR2, six predictors give the best models. Based on the related analysis in R programming BIC produced three best predictors (i.e. x2, x4, and x8), while ADJR2 led to six predictors (i.e. x1, x2, x4, x7, x8, x10).
Selection methods of predictors (a) ADJR2 information criterion (b) Bayesian information criterion
Selection methods of predictors (a) ADJR2 information criterion (b) Bayesian information criterion
To finalize our model, we employed lasso regression analysis by considering the best and bestclose lambda values, and relaxed lasso regression (Table 8). When considering the best lambda value, lasso regression identified nine predictors (i.e. x1, x2, x3, x4, x7, x8, x10, x12, and x13) with a significant effect on the outcome variable. However, the bestclose lambda value produced more robust findings, demonstrating that only four predictors (i.e. x1, x2, x4, and x8) significantly affect the outcome variable.
Lasso regression with best and bestclose lambda values
| Lambda = grid (best) | Lambda = grid (bestclose) | ||
|---|---|---|---|
| Intercept | 1.700900e-16 | Intercept | 1.43E−16 |
| x1 | −1.72E−01 | x1 | −6.56E−02 |
| x2 | 3.29E−01 | x2 | 2.52E−01 |
| x3 | −2.96E−02 | x3 | – |
| x4 | 1.47E−01 | x4 | 4.08E−02 |
| x5 | – | x5 | – |
| x6 | – | x6 | – |
| x7 | 1.32E−01 | x7 | – |
| x8 | 2.16E−01 | x8 | 1.21E−01 |
| x9 | – | x9 | – |
| x10 | −9.83E−02 | x10 | – |
| x11 | – | x11 | – |
| x12 | −3.44E−02 | x12 | – |
| x13 | 3.53E−02 | x13 | – |
Source(s): Authors’ own work
After these two lasso regressions, we employed relaxed lasso regression based on the results of lasso regression via the bestclose lambda value (see Table 9). According to these findings, the best predictors of firm performance are x1, x2, x4, and x8. Except for x1, all predictors have a positive effect on firm performance. We also visualize the lasso regression results by using the Lars Package. As shown in Figure 5, x2, x4, and x8 have positive and significant effects, while x1 (black-colored line) negatively impacts firm performance. This figure also statistically supports the lasso regression results.
Relaxed lasso regression results
| Estimate | S.E | t-statistic | p-value | |
|---|---|---|---|---|
| Intercept | 1.97 | 0.51 | 3.82 | 0 |
| X1 | −0.16 | 0.07 | −2.14 | 0.034 |
| X2 | 0.27 | 0.06 | 3.97 | 0 |
| X4 | 0.16 | 0.06 | 2.43 | 0.016 |
| X8 | 0.2 | 0.06 | 3.17 | 0.001 |
Note(s): S.E. denotes standard error
Source(s): Authors’ own work
Discussion
Despite the growing importance of market and nonmarket strategies in hospitality businesses (Köseoglu et al., 2020; O’Regan & Choe, 2019), more research is needed on their implications for hotel enterprises. This research provides academics with useful information for developing routes and validating the market strategy paradigm. This study’s findings offer several theoretical contributions to academics and researchers. The current study reinforces configuration theory by combining market and nonmarket methods as business performance indicators in hospitality firms.
Cost leadership has a considerable but unfavorable influence on hotel performance. This finding contradicts some previous research (Lee et al., 2021; Sheehan & Foss, 2007), suggesting the need for additional work. Moreover, a differentiation strategy, rather than cost leadership, is more likely to lead to increased hotel performance. Comparative research could help improve our understanding of the effects of these market strategies on hotel performance.
The observation of a highly significant effect of company size on NMS, in particular, expands past studies on the link between hotel characteristics and firm performance (Audia & Greve, 2006; Coad & Hölzl, 2009). In addition, this study’s conceptual relationship between global growth and performance suggests that hospitality organizations that adopt international expansion may improve their performance and significantly expand prior research findings (García-Almeida & Yu, 2015; Johnson & Vanetti, 2007). The discovery that nonmarket orientation significantly affects company performance clarifies the theoretical relationship between NMS and performance, both adding to public choice theory and broadening current findings on NMS-hotel performance links.
More importantly, the data show that nonmarket orientation fully mediates the relationship between relative company size and overall performance. By shedding light on how company size boosts business performance through the intervening function of the NMS, this study contributes to the hospitality management literature in a relevant and innovative manner.
Contrary to extant theory and literature (e.g. Bentzen, Madsen, & Smith, 2012; Park & Jang, 2010), the results do not support a link between business size and global growth, suggesting that additional empirical research is required. Interestingly, our analysis found no empirical evidence of a direct connection between business size and cost leadership.
More importantly, our research has empirically explored the best predictors of firm performance in hotel organizations by performing machine learning approaches. The findings revealed that cost leadership (x1), unique strategy (x2), growth orientation (x4), and CSR strategy (x8) are the variables affecting firm performance. Cost leadership negatively influenced firm performance, while three other variables significantly and positively impact hotel performance. Among the positive effects, unique strategy has the most significant impact on hotel performance and gives a fresh starting point for firm performance research in hospitality that will significantly widen or extend earlier studies.
In sum, despite the increasing importance of market and nonmarket strategies in hotel organizations, research on hotel company size, strategies, and performance is scarce. This study has provided crucial insights for academics regarding theory development and verification regarding business size, market and NMS, and overall performance in this respect. As detailed below, the conclusions of this study provide substantial theoretical contributions to the literature. First and foremost, the current study validated the NMS’s applicability to the tourism industry. Second, this study conceptually validated the assumption that company size influences NMS, which influences firm performance. Third, given the significant intervening role of NMS in this dyad, this study provides theoretical insights into market and nonmarket strategies and their involvement in the link between company size and overall performance based on configuration and public choice theories. Finally, machine learning analysis confirmed the significance of three approaches (i.e. unique strategy, growth orientation, and CSR) in improving overall performance of hotel organizations, providing avenues for future scholars interested in strategic predictors of firm performance in the hospitality industry.
Theoretical implications
In terms of theoretical contributions, the study bridges a significant gap in the literature by examining both market and nonmarket strategies simultaneously. It highlights how NMS can complement market strategies, particularly in the context of the hospitality industry, where regulatory and social environments play a crucial role (Eid et al., 2021). The findings underscore the importance of firm size as a determinant of strategic orientation. While firm size alone did not directly impact performance, it significantly influences the adoption of NMS, which in turn affects performance. This insight adds a dimension to the understanding of strategic management in the hospitality sector. Additionally, the use of both PLS-SEM and machine learning provides a comprehensive analysis of the data, allowing for the identification of both linear and nonlinear relationships. This methodological innovation enhances the robustness of the findings and offers a model for future research in strategic management.
This study advances configuration theory by illustrating how market and nonmarket strategies coalesce to form coherent strategic configurations that optimize firm performance. The results highlight that these configurations are context-dependent, with NMS acting as a key driver of performance in industries where external stakeholder engagement (e.g. political and social interactions) plays a significant role. This finding underscores the principle of equifinality, demonstrating that different strategic combinations can yield comparable performance outcomes depending on organizational context. For hospitality firms, the interplay between market and nonmarket strategies reveals a nuanced mechanism through which firms achieve competitive alignment.
The findings also provide empirical support for public choice theory by demonstrating that larger firms are more likely to adopt NMS, leveraging their resources to influence regulatory and social environments in their favor. This reinforces the notion that firms operate as rational actors, optimizing their engagement in political and social arenas to maximize returns. In particular, the mediating role of NMS highlights how these activities serve as a bridge between firm size and performance, offering new insights into the strategic calculus behind nonmarket activities.
Social exchange theory adds a relational dimension to this narrative by framing nonmarket strategies as reciprocal exchanges between firms and their stakeholders. These exchanges are characterized by a balance of costs (e.g. lobbying expenses, CSR investments) and anticipated benefits (e.g. stakeholder goodwill, favorable policies). Our findings illustrate how firms strategically engage in nonmarket activity to cultivate relational capital, which, in turn, enhances performance outcomes. Social exchange theory enhances our understanding by emphasizing the importance of trust and mutual benefits in driving the effectiveness of nonmarket engagement.
These theoretical insights collectively demonstrate that the strategic value of nonmarket strategies lies in their dual role as a performance enhancer and a contextual mediator. By integrating these perspectives, our study contributes to a more holistic understanding of how firms navigate complex strategic environments.
Practical implications
The study provides actionable insights for hospitality operators and marketers. Emphasizing global growth and nonmarket strategies, such as CSR and political engagement, can improve performance (Bonardi et al., 2006; Mellahi et al., 2016). Operators should consider these strategies to enhance their competitive advantage in a complex and dynamic market.
The mediating role of nonmarket strategies highlights their importance in the hospitality industry. Firms should consider investing in building strong relationships with stakeholders, including government entities and community organizations, to enhance their market position and performance. Stakeholder links are crucial because they build trust and facilitate collaboration, which can improve a firm’s reputation and customer loyalty. These relationships enable firms to anticipate and address stakeholder concerns more effectively, minimizing resistance to business initiatives. Additionally, strong stakeholder connections can provide valuable insights and resources, enhancing innovation and resilience in dynamic markets (Harrison & Wicks, 2013; Hillman & Keim, 2001; Mbalyohere & Lawton, 2018).
Our findings also suggest that strategic approaches should be tailored to the size of the firm. Smaller firms may need to focus more on niche markets and cost leadership, while larger firms can leverage their resources to pursue more aggressive global growth and nonmarket strategies. Indeed, larger firms generally have more resources and brand recognition, while smaller firms often benefit from niche strategies or flexible, innovation-driven approaches that exploit agility. Moreover, large firms typically engage more in nonmarket strategies to manage stakeholder relationships, while smaller firms may prioritize direct market competition due to resource constraints (Mellahi et al., 2016; Parnell et al., 2024).
Conclusion, limitations, and future research
We surveyed 142 managers and other professionals in US hospitality firms to investigate the links among firm size, market strategies, and nonmarket strategies. Firm performance was negatively associated with broad cost leadership and positively associated with a global growth emphasis. Although firm size did not influence either market strategy or performance, NMS fully mediated the relationship between firm size and performance. Our study underscores the potential value of employing both PLS-SEM and machine learning.
Our study demonstrates the significant value of machine learning techniques in strategic management research. The application of machine learning allowed us to identify key predictors of firm performance that can be overlooked when applying PLS-SEM alone. Specifically, the machine learning analysis revealed NMS had a stronger and more complex influence on performance than previously understood. The machine learning models identified that firm size positively influenced nonmarket orientation, which in turn had a significant positive impact on firm performance. Additionally, our machine learning findings showed that while broad cost leadership had a negative impact on performance, global growth orientation positively influenced firm performance, further elucidating the nuanced effects of these strategies.
While our study provides valuable insights, it also has several methodological limitations. The reliance on self-reported data from managers may introduce bias. Future research should incorporate objective performance measures to validate the findings. The cross-sectional nature of the study limits the ability to draw causal inferences. Longitudinal studies are needed to explore the dynamic relationships between firm size, strategic orientation, and performance over time. The sample size of 142 firms, while adequate for PLS-SEM, may not be representative of the broader hospitality industry. Future studies should include larger and more diverse samples to enhance generalizability.
We also acknowledge the potential for non-response and selection bias. Prolific invites qualified respondents (i.e. managers and professional working full-time in the U.S. hospitality industry) to participate in a study. Surveys typically garner responses from about one-half of the qualified respondents. We achieved 142 responses from a pool of 240 in 72 hours, a 59.1% response rate consistent with expectations. Nonetheless, non-response bias is possible, as 98 qualified individuals chose not to participate. Moreover, the Prolific pool includes only a small fraction of qualified respondents because individuals opt in as participants. The support for Prolific in academic research (see Eyal et al., 2021; Tang et al., 2022) notwithstanding, non-response and selection bias are possible in the sample.
We also identified several future research directions. First, the distinctions between market and nonmarket orientations are challenging to make in practice (Funk & Hirschman, 2017; Zhang et al., 2020). Firms often take positions on social, environmental, and political issues and engage in other nonmarket tactics to achieve market goals. Moreover, the extent to which different nonmarket approaches should be integrated into a broad NMS is unclear (Scherer, Rasche, Palazzo, & Spicer, 2016). Additional scholarship in the hospitality industry is required to identify and corroborate nonmarket strategies at the firm, strategic group, and industry levels.
Second, the short- and long-term costs of market and nonmarket strategies are intuitive, but the long-term performance effects of nonmarket approaches remain unclear (Funk & Hirschman, 2017; Mellahi et al., 2016). Nonmarket strategies can promote short-term performance but also take resources away from customers, competitors, technology, and other market considerations, potentially damaging the organization in the long run. Specifically, the extent to which NMS creates long-term benefits that justify the costs and unintended consequences is not yet known (Dorobantu et al., 2017). Moreover, a firm’s market strategy might drive financial performance, while its nonmarket strategy drives non-financial performance (Zhang et al., 2020). The link between NMS and firm performance supported herein does not resolve this dilemma because it does not address an appropriate balance between nonmarket and market strategies.
Finally, this study is based on perceptions of managers and other professionals in a single industry. A follow-up study that includes data from customers and other stakeholders would expand the literature (Alhouti et al., 2016; Yim, 2021).
