Abstract
Despite its practical relevance for internationalizing firms, organizational ambidexterity has received limited attention in international marketing research. Existing literature largely focuses on ambidexterity within a single functional domain and reports inconsistent evidence regarding its performance outcomes. The authors of this article advance knowledge by capturing contingent performance effects of exporting firms’ ambidexterity in both market and product development domains. The authors test sequential forms of ambidexterity and their effects on performance using firms’ self-reported ambidexterity measures and lagged, objective performance data. The findings reveal that market ambidexterity enhances product development ambidexterity, which subsequently contributes to return on sales (ROS) but not to sales growth. Interestingly, market ambidexterity has a direct negative effect on ROS but a positive indirect influence through product ambidexterity, indicating a suppression effect. Knowledge leverage capability strengthens the relationship between market and product ambidexterity. Additionally, the performance impact of product ambidexterity is contingent: Perceived market opportunity prevents it from eroding sales growth, while the degree of internationalization amplifies its contribution to ROS. These findings underscore the complex, domain-specific, and conditional nature of ambidexterity's performance implications in export settings.
Tushman and O’Reilly (1996) coined the term “ambidexterity” to describe firms that simultaneously exploit existing competencies in the short term and explore new, long-term opportunities. The enduring interest in ambidexterity lies in that it enables firms to manage evolutionary and revolutionary changes, but that it is difficult to use because it requires trade-offs in finite resource allocations between exploration and exploitation (Nemkova et al. 2015). Ambidexterity is especially pertinent to international marketing (IM), as internationalization is not always forward moving and monotonic (Sousa, Li, and He 2020). Firms need to understand how to gain from more and less cautious modes of growth overseas. Still, there are gaps in knowledge of the performance relevance of ambidexterity in IM.
First, prior IM ambidexterity studies (e.g., He and Wong 2004; Su et al. 2022) tend to focus on the product innovation domain and overlook the market domain (see Table 1). This omission is important as the flexibility to adopt alternative courses of action in IM is a function not only of international product transformation, but also of transformation of the strategy for markets entered (Gomes, Sousa, and Vendrell-Herrero 2020). Further, the market-selection decision can widen the firm's knowledge portfolio with valuable insights for product development (Bıçakcıoğlu-Peynirci and Morgan 2022; Jin, Zhou, and Wang 2016). Overemphasis on internal product development and inattention to external market needs has resulted in innovations remembered for underperforming overseas. For example, Uber downplayed the importance of market factors in the design of its offering in Southeast Asia, precipitating its exit from the region (Economist 2018).
Empirical Research on the Ambidexterity to Performance Relationship in International Marketing.
Notes: We exclude studies that focus on business unit, rather than organizational, ambidexterity and those with a domestic study context. DOI = degree of internationalization, ROS = return on sales.
Second, with few exceptions (e.g., Hughes et al. 2010; Yan, Tsinopoulos, and Xiong 2021), IM scholars have applied ambidexterity to foreign subsidiaries and cross-national samples rather than to exporters. Yet, in both the globalization and deglobalization eras, exporting is the low-risk entry mode offering efficacy against a backdrop of uncertainty (Chabowski et al. 2018). Exporters oversee a wide footprint of cross-border activities, and face problems with the low availability and accessibility of the export product-market information needed to make good decisions (Cadogan, Kuivalainen, and Sundqvist 2009). Managers seek ways to lower the costs of searching for information required for exporting (Iurkov, Koval, and Zaefarian 2023). Ambidexterity's knowledge creation and development properties could help with the increased levels of uncertainty inherent in export management (Cui, Walsh, and Zou 2014). It may also offer the flexibility required for exporters to successfully navigate evolutionary and revolutionary changes across several markets.
Third, while some IM studies find a positive effect of ambidexterity on performance (Geerts et al. 2018; He and Wong 2004), others show a negative and/or nonsignificant impact (Wu et al. 2020; Yan, Tsinopoulos, and Xiong 2021). This mixed evidence suggests that ambidexterity's link to performance is complex and heterogeneous. One source of complexity is the measurement of performance. Different components may yield inconsistent findings. A second source of complexity is that ambidexterity may not improve performance under all circumstances; unknown explanatory mechanisms involve boundary conditions. Robust theory building depends on IM studies’ full engagement with the sources of complexity that contribute to inconsistent performance findings.
Our study thus answers the question: Do market and product development ambidexterity sequentially lead to export performance outcomes, and under what conditions? We form domain-specific ambidexterities combining exploitation and exploration facets. Export market exploitation refers to improving and elaborating on the firm's current export market and customer knowledge, processes, and skills, whereas exploration involves obtaining and deploying new export market and customer knowledge, processes, and skills (Özsomer and Gençtürk 2003). Product development exploitation refers to refining and extending the firm's existing product development knowledge, processes, and skills, while exploration involves acquiring and harnessing new product development knowledge, processes, and skills (Yalcinkaya, Calantone, and Griffith 2007). We advance a model drawing on organizational learning and dynamic capabilities theories that is tested on a sample of Portuguese exporters.
The study contributes to the IM literature in various ways. First, our research is novel in examining market and product development ambidexterities, which reflect the two most fundamental functional domains (Voss and Voss 2013). We find that market ambidexterity is an antecedent of, and its performance relevance is mediated by, product ambidexterity, depending on the performance component. We also observe that the link between market ambidexterity and product ambidexterity is strengthened by knowledge leverage capability—the exporter's ability to leverage and use accumulated knowledge from its past experiences for new product-market opportunities. The study adds to our understanding of the use of experiential knowledge mechanisms in IM (Assadinia et al. 2019).
Second, our novel focus on exporters and their performance is apposite as this entry mode's flexibility and efficacy could be unlocked by ambidexterity (Spyropoulou et al. 2017). Further, in response to the inconsistent findings for ambidexterity's performance relevance, we take a multicomponent approach. We examine sales growth and return on sales (ROS), using lagged, objective data. Encouraged by regulatory interest in trade, exporters have shown enduring interest in fast sales growth overseas (Economist 2023). But a focus on sales growth alone ignores that growth can be at the expense of profits (Katsikeas et al. 2016). We show that neither market ambidexterity nor product development ambidexterity affects sales growth. Both impact ROS, although surprisingly, market ambidexterity has a negative effect.
Third, our findings add to select recent studies (e.g., Jin and Zhou 2021) in identifying moderators of the ambidexterity-to-performance link. We show how perceived market opportunity and degree of internationalization (DOI) influence the efficacy of product development ambidexterity in generating export performance outcomes. Perceived market opportunity refers to the tendency of the firm to interpret market changes in a positive light, whereas DOI taps the firm's scale (i.e., quantity of export business relative to overall business) and scope (i.e., range of countries and regions) of export operations (Cadogan, Kuivalainen, and Sundqvist 2009). The results show that exporters with greater perceived market opportunity avoid a negative link between product ambidexterity and sales growth, while those with greater DOI are better able to achieve ROS because of such ambidexterity.
The Domain of Ambidexterity
Our review of prior IM work on ambidexterity reveals that studies are silent on the sequential nature of the concept across distinct activities or domains, even though firms can pursue rewards from ambidexterity along various value-chain functions (Stettner and Lavie 2014). A review by Turner, Swart, and Maylor (2013) notes that the lack of a multidomain analysis of ambidexterity represents an inadequacy within existing theory. In other words, some aspects of the ambidexterity phenomenon may remain out of sight because they do not fit the assumptions of the prevailing conceptualization.
All but one of the extant IM studies on the relationship of ambidexterity to performance in Table 1 focus on the product innovation domain. The exception, which captures the market domain, is Lisboa, Skarmeas, and Lages (2013). Yet that study on exporting did not consider the product domain, ruling out testing the sequential nature of ambidexterity. In cross-border trade, product development ambidexterity requires constant scrutiny of the marketplace and the need to motivate and accomplish change quickly (Yalcinkaya, Calantone, and Griffith 2007; Yan, Tsinopoulos, and Xiong 2021). This logic implies that, for exporters, market ambidexterity antecedes product development ambidexterity.
Theoretical Background and Hypotheses
Our study views exporter ambidexterity through two complementary lenses: organizational learning theory and dynamic capabilities theory. Organizational learning theory considers how firms learn and adapt to environments (March 1991). Among the perspectives proposed to explain the ways that international firms develop and harness knowledge and how learning shapes performance, the concepts of exploitation and exploration occupy a distinctive place—as the logics needed to ensure current and future viability, respectively (Lages, Jap, and Griffith 2008). Securing ambidexterity across these very different learning activities gives firms a chance to “outrun environmental selection issues” (He and Wong 2004, p. 483), but the balance is hard to achieve. Exploration of new ideas slows the improvement of existing skills, while greater competence in current practices discourages experimentation with new ones (March 1991).
Dynamic capabilities theory posits that complex routines embedded within the firm keep a firm's resource base relevant considering external changes (Birkinshaw, Zimmermann, and Raisch 2016). The theory emphasizes processes that involve sensing and seizing opportunities to reconfigure resources to align with changing sources of competitive advantage (Pfajfar, Mitręga, and Shoham 2024). Exploitative and explorative learning are capabilities that enable firms to understand and respond to market needs faster and better than competitors, develop advantageous offerings, and realize superior performance outcomes (Vorhies, Orr, and Bush 2011). Exploitative and explorative capabilities support firms in seizing and sensing opportunities, respectively (Birkinshaw, Zimmermann, and Raisch 2016). When combined, they complement each other by forming a dynamic learning cycle (Wei, Yi, and Guo 2014). Indeed, through ambidexterity, firms can apply lessons from international operations to prevent repeating errors and use creative methods to develop new offerings and reach new foreign markets (Yalcinkaya, Calantone, and Griffith 2007). Thus, ambidexterity is a transformative, dynamic capability that enables exporters to fit the configuration of their resources with product-market settings.
Nevertheless, while firms with dynamic capabilities might attempt to link to the marketplace, the firm does not look first to learn from external developments. Dynamic capabilities may be vulnerable to an “inside-out myopia” (Day 2011, p. 187). That is, they are based on what the firm already knows it can apply to product-market opportunities (inside-out), rather than on skills and competencies that help the firm understand changes taking place in its markets (outside-in) (Morgan, Feng, and Whitler 2018). We propose that such thinking necessitates export market ambidexterity preceding product development ambidexterity. For the former the focal point is outside of the firm (external export customers and their requirements), whereas for the latter the focal point is inside (internal innovation activities and products).
Outside-in market activities encourage customer-centric conversations internally, which can support the development of new products (Quach et al. 2020). But conversations linking outside-in and inside-out capabilities require experiential learning competency in order for a firm to integrate past technical and market insights into ongoing processes (Abdi, Aulakh, and Ma 2024; Morgan, Feng, and Whitler 2018). We thus posit that the link between export market and product development ambidexterities is moderated by knowledge leverage capability.
Further, to leverage product development ambidexterity to achieve superior performance levels, firms need a way of orienting learning processes of this inside-out dynamic capability toward impending external changes. They can do so using a vigilant management approach that emphasizes the need to sense and adapt to changes (Day 2011). For example, managers might anticipate emerging customer requirements or regulatory flux in foreign markets. Thus, we posit that perceived market opportunity and DOI moderate the relationship between product development ambidexterity and performance. Perceived market opportunity is a skill that transforms apparent external noise into meaning faster. Firms that interpret market changes positively are aware of opportunities sooner and likely to act on partial information (Siegel and Renko 2012). DOI and greater exposure to, and immersion in, export markets should allow for deeper market insights and ready adaptability (Assadinia et al. 2019).
Export Market Ambidexterity and Product Development Ambidexterity
There is abundant evidence that market-related information is crucial for new product development activities generally (Quach et al. 2020) and for building product development capability specifically (Yalcinkaya, Calantone, and Griffith 2007). The stronger the firm's market knowledge base, the better its ability to recognize and pursue new product opportunities within the innovation process (Eisenhardt and Martin 2000). More specifically, when firms investigate current and new export markets, they deal with multiple and distinct customer specificities and widen their market knowledge repertoires, incorporating knowledge beyond present boundaries and enhancing the chances of unique knowledge combinations (Kim 2016). These combinations, utilizing disparate foreign-market information, ideas, and concepts, provide a particularly rich set of inputs for product development (He and Wong 2004).
New market knowledge, especially from diverse and unfamiliar foreign markets, can trigger product innovation by revealing cultural nuances, emerging customer trends, novel usage patterns, updated competitive dynamics, and unique regulatory effects (Buratti and Laukkanen 2023). Such insights, gained from exploitative and exploratory market activities, go beyond existing understanding and routines, acting as a catalyst for product innovation (Am et al. 2020). Discovering and acquiring fresh market knowledge not only informs but actively stimulates product ambidexterity by prompting firms to challenge established assumptions, reconfigure existing product offerings, and develop new ones that align with market requirements, thereby expanding the firm's innovation horizon.
Knowing the export market allows for a better understanding and forecasting of, and more comprehensive response to, export customers’ current and latent needs and expectations (Osmanoglu, Özsomer, and Biliciler 2024; Ramaswami, Srivastava, and Bhargava 2009; Sayin et al. 2024). Export customers act not only as sources of needs but also as providers of solutions-related knowledge. They may even serve as product development initiators, by requesting new features or functions (Chang and Taylor 2016). The presence of multiple exploitative and exploratory perspectives on export customer needs and solutions can bring about broader mental maps, increasing the likelihood of preventing myopic information and prompting creative resource solutions (Wang and Rafiq 2014). Hence, export market ambidexterity helps the firm avoid the trap of favoring only incremental solutions to product development problems. Further, as market ambidexterity emphasizes, in part, exploiting current customers’ needs, subsequent product development processes are unlikely to go too far in the direction of developing new skills, processes, and technologies at the expense of ongoing efforts to refine products. In sum, the combination of current and new export market knowledge can furnish the exporter with insights and synergies that facilitate the exploitative and explorative capabilities of ambidextrous product development. Thus:
We expect knowledge leverage capability to positively moderate the relationship of export market ambidexterity to product development ambidexterity. When exporters actively engage with current and new export markets, they enhance the chances of unique knowledge combinations being useful for product development (Kim 2016). However, a firm’s ability to capitalize on knowledge combinations and generate synergistic results depends on integration of the acquired market knowledge with the existing knowledge base of the firm (Raisch and Birkinshaw 2008). Learning from outside the firm involves both the acquisition of external knowledge and the ability to refine and harness it using the firm's knowledge platform (Quach et al. 2020). Knowledge leverage capability employs knowledge gained from the exporter's past experiences for new product-market opportunities. Firms with such a capability are adept at experiential learning, based on the sharing and dissemination of accumulated knowledge. For instance, ideas originating from a previous foreign customer and process alterations that happened because of them can be employed to make sense of ongoing customer insights. This integration enhances the chances of making the connections among disparate foreign-market information, ideas, and concepts that benefit product development (He and Wong 2004). Effective integration of market knowledge inside the firm can assist the firm in developing a set of incremental and radical new products that are appealing to known export markets and have the potential to open up new markets (Zhang, Wu, and Cui 2015).
An exporter with low knowledge leverage capability would lack the mechanism by which experiential knowledge is developed, maintained, and used to drive strategy development (Assadinia et al. 2019). Without the ability to harness accumulated experiences in serving export markets, the exporter is likely to struggle to understand the idiosyncrasies of different markets. It would not be able to capitalize on insights into current and new customer specificities through a product innovation strategy that embraces the complexities, and sophistication, of combining exploitation and exploration logics. New product development is likely to proceed based largely on internal technical skills and competence. Therefore:
Product Development Ambidexterity and Performance
Export performance encompasses both sales growth and ROS, reflecting key components from the product-market and financial domains, respectively (Morgan 2012). Sales growth, defined as the percentage change in sales over time, indicates the firm's ability to sustain its competitive edge overseas (Assadinia et al. 2019; Cao and Li 2015). ROS, defined as the ratio of net operating profit to sales, reflects the extent to which sales revenue is translated into profit relative to operating expenses (e.g., Kostopoulos et al. 2011). Considering both metrics enables a better understanding of how product development ambidexterity translates into performance outcomes across different facets of export success.
Blending and recombining new and current product knowledge can assist firms in growing sales revenue by enabling them to be innovative without compromising existing routines and to flexibly pursue diverse customer groups (Hsu, Lien, and Chen 2013). On the one hand, the exporter's know-how in sensing and pursuing new product opportunities would bolster its efforts to experiment with up-and-coming technologies, original features, and new product solutions (Yalcinkaya, Calantone, and Griffith 2007). This exploratory process of leveling up results in next-generation technical advances, which the exporting firm can integrate into its repertoire to open up new customer-revenue streams (Wang and Li 2008). The exporter can grow sales in export markets in which experimentation and speed are crucial considerations (He and Wong 2004). On the other hand, the exporter is likely to benefit from exploiting prior product development knowledge, which boosts the chances of competitive success in mature export markets where incremental innovations are important (Sarkees, Hulland, and Prescott 2010). Also, ambidextrous combinations of new and existing product development knowledge would produce varied insights, features, functions, and methods that are hard for rivals to disentangle and reproduce (Hill and Birkinshaw 2014; Li and Huang 2012). An exporter able to sustain its competitive edge can more easily grow sales across markets.
The capability to refine existing and acquire new product development know-how can likewise improve ROS, by enabling exporters to achieve operational efficiencies and better margins (Kostopoulos et al. 2011). On the one hand, the ability of product development exploration to level up in manufacturing, R&D, and other resource areas raises the productivity of creative innovation. Creative product innovation emphasizes adding attractive features that reduce customer uncertainties linked to the perceived newness. Firms using this strategy competently enjoy early-mover advantages and, thus, can charge premium prices to achieve more profitable sales (Lee and Zhou 2012). On the other hand, the exporter is expected to benefit from prior product development experiences by obtaining learning-curve effects and reducing the chances of costly mistakes in R&D (Li and Huang 2012). Exploitation enables the launch of new products at lower costs, releases resources for use elsewhere, and increases the chances of financial viability in stable export markets that favor incremental innovations (Sarkees, Hulland, and Prescott 2010). Finally, firms with product development ambidexterity can realize efficiency benefits by further penetrating their existing customer base—especially segments with high profit margins—to offset the costs of attracting new customers prepared to fund the next generation of innovative products (Morgan, Slotegraaf, and Vorhies 2009). In light of the preceding considerations:
We expect perceived market opportunity to positively moderate the relationship of product development ambidexterity with sales growth, specifically. Product development ambidexterity leads to sales revenue growth as it affords firms the strategic flexibility to simultaneously pursue diverse customer groups, and ambidextrous combinations of product development knowledge are hard for rivals to reproduce (Hill and Birkinshaw 2014). When perceived market opportunity is high, export managers’ tendency to interpret market changes as having positive rather than negative implications for the firm and as representing a potential gain rather than a loss helps the firm recognize and pursue short- and long-run opportunities for innovation (Day 2014). Perceived market opportunity (i.e., the mindset of transforming external noise into positive meaning faster) would encourage the exporter to follow up on opportunities to simultaneously grow sales across diverse markets and customer segments. Such entrepreneurial opportunities are not equally recognizable to all firms and can give an exporter that is adept at blending new and current product development knowledge a competitive edge with which to build foreign sales (Miocevic and Morgan 2018).
In the absence of perceived market opportunity, an exporter is expected to struggle to capitalize on changing environmental conditions by generating innovative products and meeting the needs of markets (Kostopoulos et al. 2011). In failing to be vigilant to external potentialities and nimble in attempts to capitalize on them, the exporter would pass up on valuable sales growth opportunities. Product development ambidexterity is likely to remain oriented toward inside-out learning processes that fixate on what the firm already knows it can apply to product-market opportunities. An exporter that is not receptive to external knowledge could be expected to fail to question the prevailing premises behind its internally focused knowledge. Firms may well level up their technical competence to develop creative new products without compromising existing development routines, but they are less likely to remain up to date in the marketplace and successfully grow sales (Kostopoulos et al. 2011). Accordingly:
We expect DOI to positively moderate the relationship of product development ambidexterity with ROS, specifically. Product development ambidexterity leads to superior profits from sales as it enables exporters to achieve operational efficiencies and better margins. DOI exposes the exporter to increased logistical and governance costs, given the physical distance between the firm and its markets (Christophe and Lee 2005). Exposure to heterogeneous foreign markets also requires the firm to learn and adapt to foreign laws, languages, cultures, and competitors (Cadogan, Kuivalainen, and Sundqvist 2009). Still, the costs may be dwarfed by the economies of scale and scope benefits associated with a high DOI (Kafouros et al. 2008). Exporting to a greater number of countries enables firms to test and sell the developed products in more export markets, which boosts the chances of benefiting from learning synergies and covering innovation costs, thereby enhancing ambidexterity's contribution to profits (Cassiman and Golovko 2011; Kafouros et al. 2008). An exporter with a greater DOI can create early-mover advantages in premium segments for radical innovations, and cash-cow effects in mature segments for incremental innovations, at scale. Further, because the exporter needs to account for the wider sets of circumstances of multiple markets, the higher level of knowledge integration afforded by product development ambidexterity is required to achieve operational efficiency and healthy returns (e.g., Kafouros et al. 2008).
An exporter that operates in fewer, similar markets requires lower levels of integration of exploitative and explorative capabilities in product development to ensure the efficiency of the marketing operation. Exporters with a smaller footprint derive less value from product development ambidexterity's synergistic cycle of exploitative and explorative innovations. Without the deeper market insights furnished by DOI, exporters cannot easily lower their average costs via more productive use of product development know-how by matching the firm's resource deployments with customer opportunities (Morgan, Slotegraaf, and Vorhies 2009). Further, there is a greater chance of the exporter missing key alterations within or across markets, and of developing offerings that are less well suited to the market, given the externalities. There is more uncertainty as to whether innovations will yield the sought-after returns and a greater risk of making a costly mistake. Hence:
Research Method
Sample and Data Collection Procedures
To achieve the aims of this research, we used a random sample of exporter firms in a western European country, Portugal, characterized by a small domestic market and economic sustainment via export operations (Gabinete de Estratégia e Estudos 2023). This country has an open, export-oriented economy; in the last four decades, export volume saw a sixfold increase in its contribution (Gabinete de Estratégia e Estudos 2023). The criticality of exporting for such economies is consistent with observation of the resurgence of merchandise trade globally (Okonjo-Iweala 2023). We decided on a multi-industry sample of manufacturing exporters to enhance the generalizability of findings (Navarro et al. 2010). Given their peculiarities in terms of regulations faced and international expansion patterns, we excluded exporters from the primary and tertiary sectors. Further, we used highly export-involved firms, defined as those for which export sales are at least 50% of total sales (Burton and Schlegelmilch 1987). Our objective performance measures tap overall performance rather than isolating the performance of export activities. Focusing on highly export-involved firms means that the firm's performance reflects export performance; we thus employ the term “export performance.” Firms are more likely to develop their activities (e.g., ambidexterity) with exporting (international markets) in mind. Applying these criteria, we developed an initial sampling frame of 571 firms.
We gathered data on market and product development ambidexterity, knowledge leverage capability, perceived market opportunity, country scope, technological turbulence, the year the firm started exporting, and slack resources, via an online survey. The lead researcher contacted the 571 firms by telephone to affirm their eligibility and identify key informants, defined as individuals who met knowledgeability requirements regarding the firm's knowledge development and creation processes and export operations and were willing to participate in the study (Campbell 1955). Then, we sent an email with the link to the online survey to these key informants. The survey was conducted in a single wave within a two-month period, resulting in 194 valid responses.
We obtained data on export intensity, regional scope, export performance (sales growth, ROS, past sales, and past ROS), firm size, year of the firm’s establishment, and industry from the Amadeus database. This database, accessed via institutional license, includes financial information for firms across Europe, namely 3,358 active manufacturing exporting firms from all regions of Portugal. Specifically, it contains historical data dating back 20 years, including firms’ financial statements, ownership structures, directors and contacts, financial strength indicators, stock data for listed firms, and export data. For survey data collected in 2021, the one-year lagged performance data obtained via the database consists of performance data for the financial year of 2022. This timing is consistent with standard practice in empirical research relying on firm-level financial data and ensures that performance outcomes are measured after the explanatory variables, supporting the temporal logic of the analysis and minimizing concerns about reverse causality. The merging process relied on the firm's name and tax identification number to ensure accurate alignment, which was validated through consistency checks based on industry classification, confirming that firms were correctly aligned across the two datasets.
Sample Profile
Sample firms have an average size of 130 employees and age of 28 years. On average they export to 17 countries and have been exporting for 23 years. They have an average export intensity of 84% (Min = 50%, Max = 100%). The industry patterns in the sample are in line with the country's exports; namely, textile, clothing, leather, and footwear industries (27%), chemical and rubber industries (20%), metallic and metallurgic industries (19%), and machinery and transport equipment industries (17%) are well represented (Gabinete de Estratégia e Estudos 2023).
Most respondents were chief executive officers (32%) or export managers (31%). Other respondents had other titles, such as marketing manager (12%) or general manager (10%). Respondents were experienced professionals, with an average work experience of 11 years in their position and 13 years at the firm. They showed their knowledgeability about the issues under study (M = 6.26; 1 = “very limited knowledge,” and 7 = “very substantial knowledge”) on a question included in the final part of the questionnaire.
Nonresponse and Common Method Bias
To analyze nonresponse bias, we compared early (first 25%) and late (last 25%) respondents on the number of export markets and years of export operations (Armstrong and Overton 1977). We detected no significant differences at the conventional level of .05. Further, we compared responding (n = 194) and nonresponding (n = 377) firms’ sales growth and ROS using secondary data from Amadeus. There was no statistical evidence of differences between the groups (t(569) = .000, p = 1.000 for sales growth and t(569) = −.085, p = .933 for ROS).
To minimize common method variance, we followed procedures suggested by MacKenzie and Podsakoff (2012). We used different data sources for dependent and independent variables, insofar as we collected archival performance data and survey data for the predictor variables. We also ensured a one-year temporal separation between the predictor and performance variables. In the design and layout of the survey, first, we guaranteed respondents’ anonymity and instructed them to be honest in their responses, ensuring that they understood there were no right or wrong answers. Second, the questionnaire was structured to avoid responses based on respondents’ beliefs of how our variables should relate. Third, evaluation of the variables used a range of response formats.
We ran Lindell and Whitney’s (2001) marker variable test to assess the presence of method bias in the data, based on a theoretically unrelated construct included in the survey, namely, Portugal's image abroad (1 = “very bad,” and 7 = “very good”). The average correlation between the study constructs and this variable was extremely low (average r = .043, average p-value = .597). Collectively, the procedural treatments undertaken and the analysis results suggest that method bias is not a problem in this study.
Endogeneity
To safeguard against potential endogeneity issues, various procedures and tests were performed. First, as endogeneity may occur due to simultaneity in cross-sectional designs, survey data on predictors were collected at time t, and performance involved t + 1 data. Second, although endogeneity may be due to measurement error, the results of the method bias testing indicate that this is not likely to be a problem in the present research (Sande and Ghosh 2018).
Third, we tested the endogeneity of the explanatory variables using Gaussian copulas (Park and Gupta 2012). This method enables modeling of the relationship between an endogenous variable and the regression error term using a copula. Initially, we assessed the nonnormality of the predictors of sales growth and ROS by running a Kolmogorov–Smirnov test with Lilliefors correction. Export market ambidexterity (p < .001), DOI (p < .001), and perceived market opportunity (p < .001) passed the test. However, the normality of product development ambidexterity (p = .200) was apparent, meaning that this construct was excluded from further scrutiny. 1 By using the package REndo in R, we calculated the Gaussian copulas for export market ambidexterity, DOI, and perceived market opportunity in 14 sales growth and ROS models (see Web Appendix A). Nonsignificant results for the copulas calculated suggest that endogeneity is not a problem for these variables.
Survey Instrument Development
We followed established guidelines to develop our survey instrument. First, we drew on the organizational learning and IM literatures to conceptualize the domain of, and effectively operationalize, each construct. We then conducted a series of in-depth interviews with ten practitioners with extensive knowledge of exporting firm processes to corroborate the relevance of ambidexterity and performance to the export context. They were also asked to evaluate the content and face validity of our measures and constructs (Hair et al. 2022). These procedures resulted in a draft survey instrument. Next, three academics with extensive knowledge in IM reviewed the instrument to check for clarity in its design, instructions, response formats, and items. The final stage was a pretest (n = 30; these firms only participated in the pilot) performed in identical conditions to the main study, which led to minor refinements to the instrument and its application method.
Measures
Existing measures from the literature were adapted to the research context at hand wherever possible. To do so, we used the in-depth interviews.
Ambidexterity
Following the literature, we considered exploitative and explorative capabilities to be orthogonal (He and Wong 2004). We adapted five-item measures for export market exploitation and exploration from prior research (Danneels 2008; Vorhies, Orr, and Bush 2011). We also modified our six- and eight-item scales for product development exploitation and exploration, respectively, from past research (Atuahene-Gima 2005; Danneels 2008; Yalcinkaya, Calantone, and Griffith 2007). Respondents were prompted to describe the extent to which the firm was utilizing these capabilities “over the last three years” (1 = “no extent,” and 7 = “great extent”).
Although past studies have used the combination of exploitation and exploration to measure ambidexterity, there is uncertainty over how best to develop the combination. When combining measures into a single index, information regarding the unique contribution of each individual component to the combined index may be lost (Edwards 1994). In this regard, we examined alternative combination approaches to assess the most interpretable index for combining exploitative and explorative capabilities (Edwards 1994). Following Edwards’s (1994) recommendations, we opted for the additive approach, whereby ambidexterity is computed as the sum of exploration and exploitation scores, each derived as the average of their respective multi-item measures:
Knowledge leverage capability
We obtained the three items for our new knowledge leverage capability measure, using Rossiter’s (2002) procedure. To develop and validate the scale, five groups were used: a group of directors of national/industrial associations (Group 1), three groups of export managers (Groups 2, 3, and 4), and a group of IM experts (Group 5). We generated items based on face-to-face interviews with members of Groups 1 and 2. Indeed, the interviews were instrumental to the foundation of the construct. For instance, a mold industry manager mentioned that “we take advantage of the knowledge we have in technical terms from our current customers’ industry and try to apply it in other segments.” The issue of leveraging what the firm had learned from one customer/market for other customers/markets appeared several times. The resulting pool of items was tested with Group 3. We used exploratory factor analysis (principal axis factoring with an oblique rotation) to check the scale's dimensionality (DeVellis and Thorpe 2021). We then collected data from Group 4 for confirmatory factor analysis using Amos 28. Validity was analyzed (Fornell and Larcker 1981), and one problematic item was removed. Data from Group 4 were used to assess the extent to which knowledge leverage capability behaves as it should with the related constructs, interfunctional coordination and adaptive performance (Iacobucci, Ostrom, and Grayson 1995). The correlation coefficients of knowledge leverage capability with interfunctional coordination (r = .371) and adaptive performance (r = .246) suggest that it behaves as expected. The experts of Group 5 then performed the final checks. Overall, the procedures used to develop the measure, together with the testing results, provide evidence of construct validity. The scale anchors were 1 = “strongly disagree,” and 7 = strongly agree”.
Export performance
Export performance is a complex construct, and consideration across multiple measures is desirable (Ribau, Moreira, and Raposo 2019). Here, we used sales growth and ROS. Sales growth taps performance effectiveness (Vorhies and Morgan 2003). It is a key metric used to measure channel performance in the general literature (Cao and Li 2015) and enables an exporter to assess the health of importer sales teams. We collected sales data for time t and t + 1 and calculated sales growth using the following equation:
ROS captures the efficiency with which the firm generates profitability from its sales. It is operationalized as the ratio of earnings before interest, taxes, depreciation, and amortization (EBITDA) to sales at t + 1. EBITDA reflects the firm's operating profit, excluding interest and taxes as well as noncash expenses such as depreciation and amortization. As a key “return on” metric, ROS provides insight into how much profit the firm obtains per dollar of sales (Brealey, Myers, and Allen 2020). 3 A poor ROS signals impending financial distress, while a strong ROS indicates operational efficiency.
Perceived market opportunity
We measured perceived market opportunity using a three-item, semantic differential scale modified from Atuahene-Gima (2005). The items were preceded by “Generally, market changes are perceived by the firm as ….”
DOI
We measured DOI using the percentage of total sales that result from exporting (export intensity), the number of countries the firm exports to (country scope), and the geographically distinct regions the firm exports to (regional scope). Following the literature (e.g., Cadogan, Kuivalainen, and Sundqvist 2009), scores for the three elements were standardized and weighted equally to form a single score.
Control variables
We included seven control variables: technological turbulence (rate of change in technology; items adapted from De Luca and Atuahene-Gima 2007; 1 = “strongly disagree,” and 7 = “strongly agree”), firm size (logarithm of number of full-time employees), slack resources (uncommitted resources that the firm possesses; items adapted from Atuahene-Gima 2005; 1 = “strongly disagree,” and 7 = “strongly agree”), industry (dummy variable with 0 = low and medium-low technology industry, 1 = medium-high and high technology industry; Eurostat’s [2022] classification based on technological intensity), born global (dummy variable for firms that started exporting in the year of their establishment; Knight and Cavusgil 2004), and past sales and ROS performance. As per our reference period for ambidexterity, we used average sales per assets (i.e., sales/assets) and ROS from the previous three years of the performance measure (Sousa, Martínez-López, and Coelho 2008). Table 2 shows our constructs and items.
Measures.
Secondary data (source: Amadeus database).
Notes: Export market and product development ambidexterity follow the additive approach.
Empirical Results
Validity and Reliability of Measures
We used Amos 28 software to run confirmatory factor analysis to assess the reliability and validity of the measures and employ the bootstrap procedure to address normality issues (Byrne 2016). The measurement model adequately fits the data (χ2(817) = 1,254.280, p = .000; comparative fit index [CFI] = .925; incremental fit index [IFI] = .927; Tucker–Lewis index [TLI] = .909; root mean square error of approximation [RMSEA] = .053; Bollen–Stine bootstrap [BS] p = .144) (Hair et al. 2022). The construct measures meet conventional standards for reliability and construct validity (Cheung et al. 2024). Specifically, all constructs present Cronbach's alpha and composite reliability values greater than .7. As regards convergent validity, all constructs have average variance extracted (AVE) values greater than .5, with the lowest loading being .65 (Table 3). The constructs present adequate discriminant validity as the correlations between each pair of model constructs are greater than the square root of the relevant constructs’ AVE (Fornell and Larcker 1981) and have a heterotrait–monotrait ratio below the cutoff of .85 (Henseler, Ringle, and Sarstedt 2015) (Table 3). Finally, the variance inflation factor of all independent variables was under 2.5, well below the accepted cutoff point (Hair et al. 2022).
Composite Reliability, AVE, and Correlations.
DOI is an index of three standardized variables and has a mean value of zero. Firms that are more international have values above 0; those that are less international have values below 0. Export intensity, country scope, and regional scope have mean values of 84.185, 16.624, and 4.546 and standard deviations of 12.186, 19.395, and 3.128, respectively.
Notes: All correlations p < −.140 or p > .140 are significant at the .05 level; square root of AVE is in bold on the diagonal. The heterotrait–monotrait ratio is presented in italics.
Structural Model Estimation
We tested the research hypotheses, again in Amos 28, with bootstrapping. Considering that our model includes moderating effects, we first transformed the variables involved in the interactions into single items, mean-centering them. Then, we used product-term analysis. That is, we created product terms by multiplying the predictor term by the respective moderator term (e.g., Cadogan, Kuivalainen, and Sundqvist 2009). In addition, we applied the recommendations of Little, Bovaird, and Widaman (2006) to deal with the potential for multicollinearity. Results for the baseline model, which only includes the control variables, appear in the Model 1 column; the direct effects results are in the Model 2 column; results for the model with the ambidexterity moderating effects are shown in the columns for Models 3a through 3c; and results for the final model with all the moderating effects are in the Model 4 column (see Table 4). We provide a visual overview of the tested relationships in Figure 1, which depicts the structural model.
Structural Model Linking Market Ambidexterity, Product Ambidexterity, and Export Performance.
Hypothesis Testing Results.
The p-values are based on a bias-corrected percentile method.
Notes: Model 1: χ2(61) = 102.400, p = .001; CFI = .968; IFI = .969; TLI = .936; RMSEA = .059; BS p = .141. Model 2: χ2(78) = 89.265, p = .180; CFI = .992; IFI = .992; TLI = .982; RMSEA = .027; BS p = .575. Model 3a: χ2(85) = 94.985, p = .215; CFI = .993; IFI = .993; TLI = .984; RMSEA = .025; BS p = .637. Model 3b: χ2(84) = 97.935, p = .142; CFI = .990; IFI = .991; TLI = .977; RMSEA = .029; BS p = .601. Model 3c: χ2(84) = 112.056, p = .022; CFI = .980; IFI = .981; TLI = .954; RMSEA = .042; BS p = .364. Model 4: χ2(97) = 125.588, p = .027; CFI = .980; IFI = .982; TLI = .952; RMSEA = .039; BS p = .471.
Model 1 explains 13% of the variance in product development ambidexterity, 4% of sales growth, and 12% of ROS. Model 2 explains 57% of product development ambidexterity, 8% of sales growth, and 25% of ROS. Moreover, the Model 2 findings support H1, due to the presence of a positive link between export market ambidexterity and product development ambidexterity (β = .655, p = .001). Product ambidexterity has no significant relationship with sales growth (β = −.165, p = .099), providing no support for H3a, but it is positively related to ROS (β = .597, p = .000), as per H3b. Model 3a explains 59% of product development ambidexterity, 8% of sales growth, and 25% of ROS. These findings support H2, endorsing that knowledge leverage capability positively moderates the export market ambidexterity and product development ambidexterity link (β = .132, p = .009). Figure 2 visualizes this effect, showing that the positive relationship between export market and product development ambidexterity is stronger when knowledge leverage capability is high.
Knowledge Leverage Capability Moderation on Product Development Ambidexterity.
Model 3b explains 57% of product development ambidexterity, 12% of sales growth, and 26% of ROS. The Model 3b findings support H4, that the relationship between product development ambidexterity and sales growth is stronger when perceived market opportunity is high (β = .204, p = .007). Figure 3 plots this relationship. Interestingly, it shows the presence of a negative relationship between product ambidexterity and sales growth when perceived market opportunity is low, which turns positive when perceived market opportunity is high. Using perceived market opportunity subsets (bottom 25% and top 25%) and multiple regression analyses, we found that at low levels of perceived market opportunity there is a negative link between product development ambidexterity and sales growth (β = −.429, p = .043). This path is nonsignificant when market opportunity is high.
Perceived Market Opportunity Moderation on Sales Growth.
Model 3c explains 57% of product development ambidexterity, 8% of sales growth, and 29% of ROS. The link between product development ambidexterity and ROS is stronger when DOI is high (β = .226, p = .016), in line with H5. Figure 4 confirms that the positive relationship is stronger when DOI is high. Finally, Model 4 includes all effects and explains 59% of product development ambidexterity, 12% of sales growth, and 29% of ROS. This complete model reinforces the observed hypothesis testing results.
DOI Moderation on ROS.
Supplementary Analysis
We further analyzed the research findings in two ways. First, to assess the mediation of product development ambidexterity in the relationship between export market ambidexterity (independent variable) and export performance (dependent variable), we used the bootstrapping method in Amos 28. This statistically based mediation test is considered preferable to the Sobel test and causal method of mediation testing due to its increased control over Type I error (Preacher and Hayes 2008). Given the nonsignificant results for the direct influence of both types of ambidexterity on sales growth, we only present the mediation test for ROS (see Table 5). The results indicate that export market ambidexterity is a predictor of product development ambidexterity (β = .655, p = .001), which, in turn, is a predictor of ROS (β = .597, p = .000). Moreover, results of the indirect effect based on 5,000 bootstrap samples confirm an indirect positive relationship between export market ambidexterity and ROS, mediated by product development ambidexterity (a × b = .396, bootstrap 95% CI = [.185, .564]). Surprisingly, export market ambidexterity itself is negatively related to ROS (β = −.335, p = .019). Thus, we observe partial mediation.
Mediation Bootstrapping Test Results.
Based on 5,000 bootstrap samples.
Second, we performed median and quartile post hoc analyses to better understand the significant direct and moderation effects. For the direct effects, using a median split, we found that firms with a higher export market ambidexterity achieved, on average, two times higher product development ambidexterity. Firms with a higher product development ambidexterity achieved, on average, 3.715 times higher ROS. Similarly, examining the quartiles, firms with export market ambidexterity in the upper quartile (top 25%) show product development ambidexterity 2.013 times higher than that of firms in the lower quartile (bottom 25%). Upper-quartile product development ambidexterity firms exhibit stronger ROS by 10.186 times.
To analyze the moderation effects, we divided our sample into high and low groups for knowledge leverage capability, perceived market opportunity, and DOI using a median split. We then compared the average product development ambidexterity scores across high versus low export market ambidexterity, and sales growth and ROS scores across high versus low product development ambidexterity. Firms with high knowledge leverage capability achieved 3.272 times higher product development ambidexterity when they also had high (vs. low) export market ambidexterity. Firms with high perceived market opportunity achieved 4.750 greater sales growth when they also had high (vs. low) product development ambidexterity. Firms with high DOI achieved 3.597 greater ROS when they also had high (vs. low) product development ambidexterity. We then reran the analysis, dividing our sample into high and low groups for the moderators based on the top and bottom quartiles. Firms with high knowledge leverage capability (top 25%) achieved 2.636 times higher product development ambidexterity when they also had high (top 25%) (vs. low [bottom 25%]) export market ambidexterity. Likewise, firms with high perceived market opportunity achieved 1.742 greater sales growth when they also had high (vs. low) product development ambidexterity. As for DOI, firms with high DOI achieved 1.643 greater ROS when they also had high (vs. low) product development ambidexterity.
Robustness Checks
We checked the robustness of our findings in various ways. First, our testing based on Edwards’s (1994) recommendations suggested the superiority of the additive approach to combining exploitative and explorative capabilities over alternatives. Yet the multiplicative approach is well used in prior work too. We thus used the multiplicative approach for export market and product development ambidexterities to retest our hypotheses (Jansen et al. 2009). The results replicate our findings for the direct effects and boundary conditions of the ambidexterity types (see Web Appendix C).
Second, given that performance was assessed at the firm level, we conducted a robustness check by testing the structural model using firms whose export sales represented at least 75% of total sales. As can be seen in Web Appendix D, both the direct and moderation effects that we hypothesized hold up. These findings increase confidence that our measures of firm performance can be used to capture export performance.
Third, considering our choice of ROS as a dependent variable, additional controls, such as a firm's leverage and asset liquidity, could be considered. Still, the inclusion of these additional controls did not have significant effects above and beyond the controls included in the model and did not change the model paths significantly.
Discussion and Implications
Theoretical Implications
As firms operating across markets face cultural, social, and economic differences, the market domain is potentially crucial in promoting the generation and deployment of overseas market knowledge within new product processes (Li and Huang 2012; Zhou, Wu, and Barnes 2012). However, previous IM studies have focused almost exclusively on ambidextrous learning within the product innovation functional domain (He and Wong 2004), to the detriment of understanding the foundational role of market ambidexterity. Export market and product development ambidexterities can be the aligned engines of a flexible learning and growth strategy (cf. Özsomer, Simonin, and Mandler 2023). We observe that export market ambidexterity contributes strongly to product development ambidexterity, and that this contribution is heightened in firms that possess knowledge leverage capability. Our study adds to available work (e.g., Assadinia et al. 2019) on the ability of firms that know how to export to leverage experiential learning mechanisms. Using accumulated knowledge effectively, inside the firm, can assist it in developing incremental and radical new products that are appealing to known and new export markets alike (Zhang, Wu, and Cui 2015).
The export performance and ambidexterity literatures are disconnected, with few exceptions (e.g., Hughes et al. 2010; Yan, Tsinopoulos, and Xiong 2021). Yet exporting firms operating across market settings encounter heightened levels of dynamism. In such settings, ambidexterity may serve as a dynamic capability that is crucial in understanding which resources and capabilities to develop and how to leverage them to achieve performance outcomes (Cui, Walsh, and Zou 2014). Nonetheless, we provide the insight that exporter ambidexterity's performance relevance depends on the type of performance. The results show that ambidexterity has an impact on efficiency (ROS) rather than effectiveness (sales growth).
Learning in product development can endow an exporting firm with the technical competencies to create enhanced and innovative products that generate operational efficiencies and healthy profits. We observe that product development ambidexterity boosts ROS. However, surprisingly, export market ambidexterity has a negative effect on ROS. This finding, while inconsistent with the organizational learning literature that suggests that forms of ambidexterity should help a firm achieve superior performance (March 1991), is in line with studies that alert firms to the practical difficulties of dealing with both exploitation and exploration (Wu et al. 2020). Reconciling existing and new export-market knowledge appears to be problematic for ROS, given the efforts and coordination costs needed (Rawley 2010). The direct negative effect of market ambidexterity on ROS closely mirrors its indirect positive effect via product ambidexterity, suggesting a suppression effect. While market ambidexterity alone may hurt ROS due to internal strain or complexity, it enables better market sensing. The export market insights obtained can help firms align offerings to market needs and fine-tune product strategies, thus enabling them to regain lost ground on ROS. The presence of such trade-offs involved in pursuing export market ambidexterity suggests the need to scrutinize specific moderating factors that could shape the balance between the opposing forces.
The results reveal that neither product development ambidexterity nor export market ambidexterity drives effective sales growth. We observe a modest R2 for sales growth. This is a surprise given managerial reports that firms’ top drivers of export growth, in the face of substantial global uncertainty, include building demand through core operations as well as expanding operations (Economist 2023). The explanation could be that sales growth is a metric by which the health of a sales team may be gauged. In export channels, firms often rely on the promotion efforts of importing distributors, irrespective of exporters’ strategies (Obadia and Vida 2024). Further, ambidexterity across and for export markets could make it difficult to establish and maintain brand awareness and, thus, to differentiate products from those of rivals. Differentiation from competitors in ways that add value for customers can lessen customer churn. Such churn reduces sales growth (Morgan, Slotegraaf, and Vorhies 2009).
Considering the intricacy and volatility of the IM settings in which ambidexterity may ensure an internationalizing firm's viability, one should consider the firm's perceived market opportunity tendency and DOI, which may act as contingent conditions. We reveal that firms that tend to perceive the market situation as having positive rather than negative implications possess a competitive edge that helps them arrest product development ambidexterity's slide toward having a negative impact on sales growth (cf. Atuahene-Gima 2005; Miocevic and Morgan 2018). Interestingly, when managers perceive low market opportunity, efforts to pursue both exploratory and exploitative product development may create internal tensions and/or resource misalignments that hinder sales growth. Further, our findings reveal that DOI facilitates the conversion of product development ambidexterity into superior ROS. Exporters that internationalize are in touch with multiple markets and customers’ preferences and needs, which helps them understand how to efficiently integrate product development exploitative and explorative capabilities to advance the contribution to profits (Kotabe, Srinivasan, and Aulakh 2002). Our findings thus add to select recent work (Jin and Zhou 2021; Wu et al. 2020) by identifying moderators of the effects of ambidexterity on the components of export performance. Moreover, the component-specific, contingent performance effects identified in this study advance the ambidexterity literature by clarifying the contextual factors that shape ambidexterity's efficiency and effectiveness, and that explain the mixed ambidexterity-to-performance outcomes reported in previous research.
Managerial Implications
Managers engaged in complex export settings can strengthen their ability to leverage ambidexterity productively by viewing export market ambidexterity as a driver of product development ambidexterity. When deciding how to allocate the firm's resources across ambidexterity activities, export managers need to initially exploit and explore knowledge about the export markets in which the firm currently operates or wants to enter with a view to integrating market knowledge with technical knowledge. Even though technical knowledge is valuable in terms of performance, export managers should capitalize on market knowledge for innovative product development. The starting point is external to the firm, grounded in a thorough understanding of export customers and market requirements. Further, firms with stronger knowledge leverage capability are better able to translate market ambidexterity into product ambidexterity. Exporters with weak knowledge leverage capability risk underutilizing market insights, thereby limiting their product development potential. It is thus important that exporting firms develop an internal, experiential knowledge platform to interpret and transfer external market insights (e.g., untapped segments and preferences for design) for the benefit of work to refine existing products and develop new ones adapted to dissimilar or similar foreign customer needs.
Our research directs export managers’ attention toward circumstances that play a role in the relationship of product development ambidexterity to export performance. Export managers should bear in mind that the impact of product development ambidexterity on sales growth is contingent on perceived market opportunity. More specifically, firms need to bolster their perceptions of market opportunity, as such a tendency can arrest a potential negative relationship between product development ambidexterity and sales growth. It follows that exporting firms lacking the vigilant mindset of transforming external noise into positive meaning and adapting their actions accordingly could benefit from a more circumspect strategy. That is, if they plan to actively pursue product exploration and exploitation to capture evolving demand and drive sales expansion, they might select markets that are easier to perceive as opportunity-rich or having upside potential.
Highly internationalized exporters are likely to achieve superior ROS from investing in ambidextrous product development strategies, compared with those with lower levels of internationalization. Firms with high international exposure should therefore align their R&D efforts with their international market reach to demand and emphasize investments in ambidextrous product development processes. In doing so, exporters can better leverage the gains associated with DOI and drive ROS across global markets. Collectively, our moderation thesis suggests that success in export markets is broadly contingent on managers’ ability to align market-related ambidextrous strategies with the firm's knowledge capabilities and to align product-related ambidextrous strategies with market perceptions and the firm's internationalization profile.
Although we observe (in Table 4) that sales growth is positively related to ROS, managers should note that achieving both performance outcomes using ambidexterity is not a straightforward matter. Our findings (e.g., the median and quartile supplementary analysis) suggest that product development ambidexterity is a much better predictor of ROS than it is of sales growth. Further, although export market ambidexterity is negatively related to ROS, market ambidexterity's contribution to ROS can be made productive via product development ambidexterity (i.e., positive indirect effect). It is not the possession and combination of market knowledge, per se, that benefits ROS. Rather, exporting firms need to use the insights provided by this knowledge in product development to generate profits. It is important that managers invest in building up their firm's knowledge leverage capability and DOI, given that these moderators have the potential to strengthen the paths of the positive indirect effect.
Limitations and Future Research
Although this study offers a set of theoretical and managerial insights, it is subject to limitations that point to directions for future research. First, because we conducted the study in the context of a single country's exporting manufacturers, care should be taken when generalizing the results of this study to other export settings. It is important not only to extend this research to other exporters, but also to compare the market and product development learning processes of exporters with those of subsidiaries, joint ventures, and strategic alliances in international markets. Second, the study neglects the role of importers. An interesting extension would be to investigate the role of importers’ knowledge transfers in fostering exporters’ market ambidexterity and, in turn, their ability to enhance and benefit from product development ambidexterity. Indeed, investigating additional capabilities and activities that facilitate the transfer and integration of external knowledge (e.g., absorptive capacity or an interfunctional coordination system) could also serve to strengthen the identified links.
Third, our sequential analysis of ambidexterity builds on the premise that market-related information assists new product development (Yalcinkaya, Calantone, and Griffith 2007). Yet it is possible to question this ordering (Miocevic, Kadic-Maglajlic, and Arslanagic-Kalajdzic 2025). Given the merit of the alternative sequence, it can be explored in future research. Relatedly, future research may provide a more holistic understanding of how ambidexterity operates and contributes to performance by adopting multilevel, multidomain designs. In this vein, it would be promising to examine how ambidexterity unfolds and interacts across organizational levels such as teams, business units, and the firm as a whole, as well as across functional domains including marketing, operations, and R&D.
Fourth, whereas the proposed model accounted for time, namely in the development of capabilities and in the lagged dependent variables, DOI was measured at one moment in time. Still, historical DOI with different time lags might play a varying role in moderating product development ambidexterity's impact on generation of ROS. Future research should investigate this temporal influence. Fifth, it would be desirable for scholars to target identifying unknown explanatory mechanisms involving boundary conditions under which an exporter's product development ambidexterity can better impact sales growth and allied components of export performance effectiveness. Finally, in the highly export-involved firms in our sample, exports account for a substantial share of total revenues and growth, making export performance closely aligned with overall firm performance. Future research could extend our approach by incorporating export-specific indicators (e.g., export sales growth and export profitability) to more precisely capture the effects of ambidexterity in export markets.
Supplemental Material
sj-pdf-1-jig-10.1177_1069031X251408340 - Supplemental material for Juggling Markets and Products: How Does Ambidexterity Impact Export Performance?
Supplemental material, sj-pdf-1-jig-10.1177_1069031X251408340 for Juggling Markets and Products: How Does Ambidexterity Impact Export Performance? by Ana Catarina Cadima Lisboa, Dionysis Skarmeas, and Matthew J. Robson in Journal of International Marketing
Footnotes
Editor
Ayşegül Özsomer
Associate Editor
Babu John Mariadoss
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
Data Availability
Data that support the findings of this article are not publicly available due to ethics requirements but are available from the lead author on reasonable request. Further restrictions apply to the availability of objective data, which were used under license and thus are not publicly available. These data are available from the lead author on reasonable request and with permission of Moody's Analytics.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
