The multifaceted role of digital transformation in restaurant resilience

Organizational resilience is commonly defined as a multidimensional concept that encompasses operational stability to withstand disruptions and flexibility to recover quickly (DesJardine et al., 2019). Among various factors, digital transformation (DT) has emerged as a critical driver to enhance organizational resilience (He et al., 2023), as it enhances firm’s adaptability to changing environments and capability to utilize accurate information (Ullah et al., 2025). While previous studies have related DT to resilience, findings remain inconsistent. Some studies argued that DT strengthens organization’s ability to adapt, innovate and respond to crisis by strengthening innovation capabilities (e.g. He et al., 2023; Browder et al., 2024), others concluded that its benefits are negated by implementation challenges (e.g. Favoretto et al., 2022), and adverse organizational impacts (e.g. Liu et al., 2024).

One reason for this inconsistency may stem from different approaches in measuring resilience (Hall et al., 2023). For example, Abidi et al. (2023) operationalized resilience as a firm’s ability to maintain demand during disruption, Ortiz‐de‐Mandojana and Bansal (2016) measured resilience through firms’ capacity to sustain profitability via cost efficiency, and Lins et al. (2017) operationalized resilience based on stock value stability driven by market confidence. However, organizational resilience reflects diverse facets of firm performance (He et al., 2023), and thus inherently a multidimensional (Hillmann and Guenther, 2021). This suggests that single metric only captures isolated dimension of resilience and contribute to inconsistent findings. This pattern underscores the need for a multidimensional measurement approach to more accurately reflect the complexity of organizational resilience.

Second, these inconsistencies stem from a lack of comprehensive framework that explains how DT enhances organizational resilience. Past studies often relied on dynamic capabilities theory to understand the DT–resilience relationship (e.g., Browder et al., 2024; Ullah et al., 2025). These studies highlighted DT’s role in fostering agility and innovative capacity to enable fast recovery during times of crisis (Browder et al., 2024). However, this approach often overlooks the role of DT in minimizing disruption by leveraging existing resources to continue operation and improve efficiency during the initial shock (Abidi et al., 2023; Liu et al., 2022). Given that resilience requires both stability and flexibility (DesJardine et al., 2019), a theoretical framework that omits this dual role offers an incomplete understanding of the DT-resilience relationship. Therefore, it is essential to consider both stability and flexibility in understanding how DT contributes to organizational resilience.

Finally, the inconsistency may be attributed to disregarding the functional heterogeneity of DT within organizations. As noted by Cavusoglu (2019), the impact of DT varies across operational areas depending on the role and the function of the technologies used. For instance, some DT initiatives focus on automating existing activities to improve production efficiency, while others use technologies to transform service delivery and improve customer experience (Cheng et al., 2023). However, past studies overlooked the heterogeneity across DT categories (e.g. He et al., 2023; Liu et al., 2022). Because DT fulfills distinct roles within the business environment (Alt, 2021), evaluating the effectiveness of the DT without considering functional heterogeneity would seriously impede the understanding of the DT–resilience relationship.

To address these gaps, this study examines the DT–resilience relationship in the restaurant industry, using the COVID-19 pandemic as the empirical context for testing this relationship. Specifically, this study first considers multiple domains of organizational resilience by investigating revenue, profitability and firm value, which represent the core areas most disrupted during the pandemic (Yost et al., 2021). Resilience in sales revenue reflects the ability to sustain sales and customer engagement during crisis (Yang et al., 2021), while resilience in profitability captures the firm’s ability to manage cost and sustains margin despite operational challenges (Ekici et al., 2025). Resilience in firm value represents investor and market confidence during crisis (Lee et al., 2024). Evaluating these distinct domains moves beyond single-metric approach to reveal whether DT creates balanced benefits across revenue generation, cost efficiency and market valuation, or if its impact concentrated in specific financial domains (He et al., 2023).

Second, this study employs the ambidexterity perspective to explain the DT-resilience relationship within the restaurant industry. Past studies have primarily relied on dynamic capabilities theory which underscores agility and innovative (flexible) capacity in achieving quick recovery (e.g. Browder et al., 2024; Faro et al., 2024). Meanwhile, the ambidexterity perspective (Raisch et al., 2009) further incorporates firm’s ability to maintain operational stability. Building on this framework, this study conceptualizes DT as a dual-function capability:

• exploitive function to achieve stability during initial disruptions (Abidi et al., 2023; Liu et al., 2022); and

• exploratory function to enhance flexibility in the post-shock recovery (Browder et al., 2024; Ullah et al., 2025).

Thus, ambidexterity allows for a more detailed understanding of the DT–resilience relationship by revealing the distinct role of DT in fostering both stability and flexibility.

Moreover, this study embeds ambidexterity framework within the two functional domains of restaurant operations: Front-of-House (FoH) and Back-of-House (BoH) to account for the industry’s functional heterogeneity. This operational distinction based on functional areas is well established in the literature (e.g. Kim et al., 2025). Consistent with the literature, this study adopts a classification that distinguishes DT according to these functional areas (Alt, 2021; Cavusoglu, 2019). FoH DT (e.g. online ordering, self-service kiosks and contactless payment) is closely linked to revenue resilience, as it enables restaurants to preserve sales channels during operational disruptions (Yang et al., 2021). By contrast, BoH DT (e.g. automated inventory, labor scheduling and enterprise resource planning (ERP) systems) supports profitability resilience by enhancing efficiency and cost management (Ekici et al., 2025). Thus, understanding how ambidextrous DT–resilience mechanisms operate within these distinct domains is essential for explaining why DT generates different resilience outcomes.

In sum, this study examines the relationship between DT and organizational resilience in the restaurant industry. Specifically, the objectives of this study are to conduct an in-depth analysis of three key relationships:

• the relationship between FoH DT and revenue resilience;

• the relationship between BoH DT and profitability resilience; and

• the relationship between overall DT and firm value resilience during the pandemic.

Organizational resilience refers to an organization’s ability to withstand stress and bounce back from a crisis (Hillmann and Guenther, 2021). Despite diverse components of resilience, past studies commonly identified two attributes in conceptualizing organizational resilience: stability and flexibility (e.g. DesJardine et al., 2019; Hillmann and Guenther (2021). Stability is an organizational ability to withstand disruption and avoid substantial decline in performance (Salwan and Gada, 2018). This means firms with stability can absorb shocks and withstand external disturbances. As a result, stable firms could minimize the adverse impact of the disruption on their performance (Abidi et al., 2023). Meanwhile, flexibility refers to an organization’s capacity to adapt and reconfigure its operation in response to crisis. Following the disruption, flexible firms have agility to adapt their operations to match new environmental demands (Ghazi et al., 2024). This adaptability allows firms to generate and evaluate alternative options and reconfigure existing practices under uncertainty (Browder et al., 2024). Therefore, flexibility enables organizations to accelerate the pace of operational and market recovery after a shock (DesJardine et al., 2019). Accordingly, this study defines organizational resilience as stability to withstand drops; and flexibility to bounce back after disruption.

DT is a fundamental change process that leverages digital technologies to redefine business models, operational processes and stakeholder interactions (Fitzgerald et al., 2014). Unlike a simple technology adoption, DT involves comprehensive organizational conversion that transforms traditional structures and fosters innovation. For example, artificial intelligence (AI) and robotic technologies have fundamentally changed the service delivery process and customer experience (Chi et al., 2020). Technologies such as Internet of Things (IoT) and machine learning algorithms enabled seamless personalized customer experiences (Buhalis et al., 2023). In the restaurant industry, DT is integrated into operational domains: FoH and BoH (Alt, 2021). FoH comprises all customer-interactive spaces including both physical and digital touchpoints (Kim et al., 2021), while BoH includes back-end operations and administrative tasks (Kim et al., 2025). Given the distinct roles and responsibilities of each area, different DT is implemented to support their unique operational needs (Chi et al., 2020). Figure 1 shows the overview of restaurant process (Alt, 2021).

Raisch et al. (2009) noted that ambidexterity is an organization’s ability to pursue exploitation and exploration simultaneously that are distinctively essential to achieving competitive advantage. Exploitation focuses on leveraging existing capabilities to improve efficiency and short-term organizational performance through continuous improvement of existing products, process optimization and knowledge integration (Benner and Tushman, 2003). In contrast, exploration is a search process that extends existing knowledge and pursuing novel opportunities via experimentation, innovation and acquisition of new abilities (Raisch et al., 2009).

In this context, the ambidexterity perspective could be applied to explain DT–resilience relationship (He et al., 2023). From a resilience perspective, organizational survival during crisis often hinges on both stability and flexibility (DesJardine et al., 2019). Exploitation-oriented DT initiatives such as automating existing process and optimization of customer service platforms can enhance operational efficiency and sustain operation during crisis (Yang et al., 2021). These technologies help stabilize performance by reducing disruptions to core activities and preserving continuity (Kim et al., 2021). Meanwhile, exploration-oriented DT such as experimenting with new business models or novel digital services can flexibly adapt to environmental disruptions and capture new opportunities that arise in the postpandemic uncertainty (Cavusoglu, 2019). These exploratory efforts provide the basis for reconfiguring resources and renewing offerings as external conditions evolve (Bilgihan and Ricci, 2024). By fostering both dimensions, DT-driven ambidexterity may equip firms with the dual capacity to absorb shocks in the short term while preparing for longer-term transformation (He et al., 2023).

Building on this perspective, this study uses ambidextrous perspective to explain the relationship between DT and resilience. Specifically, DT enables resilience by equipping firms with both the stability to absorb shocks (through exploitation) and the flexibility and creativity to adjust to sudden changes (through exploration).

This study argues that FoH DT contributes to revenue resilience by leveraging exploitation (stability) and exploration (flexibility) in the customer-facing domain. Because FoH is where demand is captured (Alt, 2021), FoH DT is related to how restaurants stabilize incoming revenue flows during disruption and regenerate demand during recovery (Kim et al., 2021). Exploitation focuses on leveraging existing resources and capabilities to stabilize existing operation during the disruption (Benner and Tushman, 2003). For instance, the use of FoH technologies such as online/mobile ordering systems, kiosks and contactless payment platforms are rapidly accelerated during pandemic (Yang et al., 2021). This facilitates restaurants’ rapid transition to delivery service model (Chi et al., 2020) and ultimately enables restaurants to offset the dine-in revenue decline and survive during pandemic (Kim et al., 2021). As such, FoH DT activates exploitation that preserves operational stability amid pandemic disruptions (Yang et al., 2021).

In addition, FoH DT also enables flexible and adaptive engagement with shifting customer needs (Bilgihan and Ricci, 2024). As noted, exploration extends existing knowledge and novel opportunities (Jansen et al., 2009). Exploration-focused FoH technologies such as IoT-based customer relationship management (CRM), dynamic promotions and AI-driven predictive analytics (Alt, 2021) create new information flows that support experimentation with new customer segments, service channels and menu bundles (Buhalis et al., 2023). Because restaurants can flexibly adjust their offerings, FoH DT help restaurants create or recapture demand in the post-shock period (Bilgihan and Ricci, 2024). This suggests that FoH DT supports restaurants to explore new business opportunities aligned with environmental change (Cheng et al., 2023), and ultimately accelerates a substantial postpandemic revenue rebound. As such, FoH DT also activates exploration capabilities that fosters flexibility in the postpandemic period (Yang et al., 2021).

In sum, FoH DT functions as dual-capability factor that contributes to revenue resilience through two distinctive mechanisms:

1. stability by exploiting existing resources to make current operation more stable; and

2. flexibility by exploring new approach that allow firms to adapt offerings in response to evolving market situations.

As FoH DT primarily functions to preserve operational security and expanding new revenue streams (Buhalis et al., 2023), this study proposes the following hypotheses:

This study also argues that BoH DT enhances profitability resilience by activating exploitation and exploration within the cost- and operations-oriented domain. BoH refers to the internal operational areas of a restaurant responsible for food preparation, inventory management and internal operations coordination (Kim et al., 2025). Therefore, BoH technologies can shape how effectively restaurants control costs during disruption and reoptimize operations during recovery.

As noted, BoH DT can stabilize existing operation by reducing variability and increasing the efficiency of core production activities (Law et al., 2025). BoH technologies encompass automation technologies (i.e. robotic kitchen assistants, kitchen display systems) and AI-supported systems (i.e. inventory management, demand forecasting and employee scheduling) that automate/reconfigure operations (Cavusoglu, 2019). These technologies help restaurants reduce labor intensity as well as streamline back-office workflows (Berezina et al., 2019), and ultimately enhance cost efficiency during crises (Law et al., 2025). That is, BoH DT enables restaurants to optimize (exploit) operational resources including labor, inventory and time to stabilize profitability during disruption (Berezina et al., 2019). As such, BoH DT contributes to maintaining stability amid pandemic disruptions (Alt, 2021).

In addition, BoH DT enables adaptive cost management and operational innovation during the recovery period (Ekici et al., 2025). Technologies such as ERP system, pricing tools, digital loyalty platforms and inventory management systems enable restaurants to quickly respond to supply–demand fluctuations and make complex pricing, promotions and inventory decision. This suggests that restaurants can generate new information flows related to pricing, promotions and inventory decisions (Bujalance-López et al., 2025). These exploration capabilities allow restaurants to rapidly reconfigure internal processes (Berezina et al., 2019), thereby minimizing inefficiencies in the operation after the crisis. This indicates that BoH DT enables restaurants to generate (explore) data-driven operational innovations that optimize resource use (Berezina et al., 2019). As such, BoH DT helps restaurants obtain operational flexibility in the postpandemic period (Yang et al., 2021).

In sum, BoH DT facilitates stability by exploiting existing operational resources to maintain cost efficiency during disruption and flexibility by exploring data-driven operational innovations to optimize resource use under changing market conditions. As BoH digital technologies have been predominantly associated with cost efficiency and profitability improvements (Alt, 2021; Cavusoglu, 2019), the following hypotheses are proposed:

Following the outbreak of the pandemic, firms often face significant uncertainty and information asymmetry regarding their present condition and future outlooks (Jindra and Moeller, 2020). Such uncertainty leads to investor caution resulting in volatility and declines in firm value as investors are limited to making informed decisions (Zhang et al., 2003). Nonetheless, past studies suggest that firm value resilience depends on firm-specific attributes that mitigate uncertainty and maintain investor confidence (e.g. Lins et al., 2017; Singal, 2015). For instance, Singal (2015) found that financial structure may mitigate or amplify value losses, while Lins et al. (2017) argued that firms with higher social capital experienced more stable stock values during financial crisis. These suggest that internal resources signal the stability and preparedness of a firm, allowing firms to protect firm value during crisis.

This study posits that a firm’s DT initiatives signal to external investors on both firm’s stability (exploitation) and its flexibility (exploration) in assessing firm value. As strategic internal resources (Hussain and Malik, 2022), DT enables firms to both withstand disruptions and adapt to evolving market conditions (Bujalance-López et al., 2025). As noted, DT-oriented firms are better positioned to exploit existing operational strengths while exploring new opportunities for adaptation and growth in changing environments (Liu et al., 2023). In the restaurant context, FoH DT enables restaurant firms to sustain revenue streams and customer engagement during disruptions (He et al., 2023), whereas BoH DT contributes to profitability resilience by enhancing cost efficiency during and after pandemic (Alt, 2021). These signals readiness and innovation-driven resilience across both stability and flexibility dimensions (Moker et al., 2020). In return, it generates investor confidence, reduces stock value shocks and ultimately leads to faster recovery during pandemic (Moker et al., 2020).

The signaling effect is particularly prominent in the restaurant context as the industry is more vulnerable to market volatility and pandemic risk (Liu et al., 2022). Moreover, restaurant industry suffers from thin profit margins and is particularly susceptible to heightened investor skepticism during the disruption (Hussain and Malik, 2022). Lins et al. (2017) notes that the signaling effect of credible resources on firm value recovery becomes more pronounced when institutional trust is generally low. This suggests that in context where uncertainty is elevated and investor confidence is weak as in the restaurant sector during the pandemic, the clarity of DT initiatives play an important role in shaping market perceptions (Fitzgerald et al., 2014). In this context, restaurant DT operates as a strategic asset that communicates resilience and future orientation to the market. Hence, this study proposes the following hypothesis:

This study analyzes quarterly data on publicly traded US restaurant companies from 2000 to 2022. Data from 2000 to 2020 are used to evaluate performance decline (H1/H3/H5), while the data from 2020 to 2021 are used to assess post-shock recovery (H2/H4/H6). Moreover, data from 2020 to 2022 are used to capture the time required to return to prepandemic performance. Financial data is collected from the COMPUSTAT database, and DT information is retrieved from 10-K filings. After excluding data with missing values, the sample includes 1,173 observations from 43 restaurant companies. Because DT adoption is not exclusive to specific service formats (Huber et al., 2010), this study includes both limited- (16) and full-service restaurants (27) in the estimation (see  Appendix in supplementary materials).

The main objective of this research is to examine the effect of DT on organizational resilience reflected in key performance indicators such as revenue, profitability and firm value. Since the concept of organizational resilience involves two dimensions of organizational stability and flexibility: stability indicates the severity of loss immediately after the shock (Salwan and Gada, 2018); while flexibility represents extent of bounce back after the disruption (Ortiz‐de‐Mandojana and Bansal, 2016), this study applies two dimensions of stability and flexibility to measure organizational resilience across revenue, profitability and firm value.

To quantify stability, this study follows prior studies (DesJardine et al., 2019) and identifies the lowest quarter for each outcome variable during the pandemic. It is important to note that each outcome variable reached its lowest point at a different time during the pandemic. Figure 2 represents the quarterly trends in revenue, operating margin and stock prices of restaurant firms from 2019Q1 to 2021Q4. Consistent with Cheema‐Fox et al. (2021), the drop period is defined as 2020Q1 for firm value and 2020Q2 for revenue and profitability. The severity of loss is measured as the percentage decline between prepandemic baseline and the lowest observed point during the pandemic. To ensure that performance drops are represented as positive values, the results are multiplied by −1. Following formulas are applied to measure the stability of revenue, profitability and firm value, respectively:

Meanwhile, flexibility is measured by the extent to which these firm performances are recovered within one year after the lowest point. Because flexibility is an ability to bounce back to normal (DesJardine et al., 2019), the one-year recovery window is consistent with the 2021 deregulation of dining restrictions by which restaurant operations began to normalize (Onedine, 2021). Accordingly, the rebound period for revenue and profitability is defined as spanning from 2020Q3 to 2021Q2, while the rebound period for firm value extends from 2020Q2 to 2021Q1. The degree of rebound is calculated as the percentage change from the lowest point during the recovery period as follows:

This study uses three independent variables to measure FoH DT, BoH DT and overall DT in the restaurant industry. To quantify these DTs, this study follows Li et al. (2022) and extracts DT-related contents from firm’s 10-K reports. Because DT has become a critical factor in achieving competitive advantage (Fitzgerald et al., 2014), firms often disclose information about their DT initiatives in their official filings such as 10-K, 10-Q reports to signal strategic commitment to external stakeholders (Li et al., 2022). Indeed, the use of 10-K reports to extract DT information has been used in the field of strategic management, finance and digital innovation in recent years (e.g. Chen and Srinivasan, 2024; Li et al., 2022). Therefore, this study measures the level of DT by analyzing the frequency of DT-related terms in firm’s 10-K reports using following process. First, this study follows Cavusoglu (2019) and created a bag-of-words that represents DT in the restaurant industry. The Cavusoglu (2019) classification has been widely adopted in the field of hospitality research as it helps identifying the distinct operational domains of restaurant technologies (Alt, 2021). Table 1 presents the DT-related keywords used in this study. Then, these keywords are systematically processed and counted through restaurant firms’ 10-K reports. Finally, the final DT levels are computed based on the number of DT-related words in firms’ 10-K reports.

This study incorporates a set of control variables in the models. Debt leverage (DEBT) is included in the model to control for the effect on firm performance (Stulz and Johnson, 1985). To account for operational differences, a dummy variable for restaurant type (LIMITED) is included where value of 1 represents limited-service restaurants and 0 represents full-service restaurants. Firm assets (ASSET) are included to control for its influence on operational efficiency and profitability. In addition, a dummy variable of Tobin’s Q (DQ) is also included to control the impact of a firm’s intangible capital (Marrocu et al., 2012). Operating cash flow (CASHFLOW) is used to control for its impact on revenue and profitability. Moreover, market share (MKT_SHARE) is included in the models to control for its contribution to the profitability (Buzzell et al., 1975). To account for financial market reactions, this study includes a dummy variable for dividend payments (D_DIVIDEND) (Sheel and Zhong, 2005). Similarly, share repurchase activity (REP) is controlled, as stock buybacks are commonly viewed as signals of undervalued equity, potentially boosting stock prices (Zhang, 2005). In addition, year dummies are included to account for unobserved time-fixed effects. Table 2 shows the variables used in this study.

To estimate the impact of DT on organizational resilience in the restaurant industry, this study uses two econometric approaches. First, a difference-in-differences (DID) model is used to examine the stability dimension to assess how exploitativeDT capabilities help firms mitigate abrupt performance drop during pandemic shock (H1/H3/H5). The DID specification captures the differential effect of the shock on firms with varying levels of DT intensity through the interaction between DT and the drop-period indicator. (Abidi et al., 2023) The estimation models are specified as follows:

where $REV_DROPi,$$MARGIN_DROPi$ and $S.PRICE_DROPi$ indicate the spread in the performance between the prepandemic baseline and the lowest performance point; $FDTit$⁠, $BDTit$ and $DTit$ are the level of FoH DT, BoH DT and overall DT, respectively; $DROP_PRDt$ represents the drop period; $DTit×DROP_PRDt$ evaluates the DID effect; $X$ is a set of control variables included in the model; and $γit$ denotes firm-quarter dummies to control for seasonality and firm-specific characteristics that remain constant over time.

Second, this study uses two-way panel estimation to capture the flexibility dimension of organizational resilience. This is to analyze how explorativeDT capabilities facilitate performance recovery from the trough caused by the pandemic shock (H2/H4/H6). Unlike performance decline, recovery takes place entirely in the post-shock period (Salwan and Gada, 2018). Because performance recovery is not estimated through time-based comparisons, this study adopts a two-way panel estimation (DesJardine et al., 2019). It is important to note that robust standard errors are used to achieve a heteroskedasticity-consistent estimator. Moreover, the Hausman specification test reveals that random-effects estimation is appropriate for the models. The estimation models for organizational flexibility are formulated as follows:

where $REV_REBit,$$MARGIN_REBit$ and $S.PRICE_REBit$ indicate the performance rebound from the lowest performance point.

Table 3 presents a descriptive statistic of variables used in this study. On average, firms with higher levels of DT adoption report greater revenue, asset size and profitability than low DT firms. These patterns are consistent across both FoH and BoH DT. Larger firms tend to adopt more DT, which may reflect their greater access to capital, skilled labor and digital skills. In addition, high DT firms maintain higher debt levels, although they also display stronger profitability and liquidity. These descriptive patterns suggest that high DT firms are generally larger and financially more robust to manage investment risks.

Figure 3 illustrates the revenue index trends by FoH DT level from the 2020Q1 to the 2021Q1. As evident, both high and low FoH DT groups experienced a sharp decline in revenue during 2020Q2. It should be noted that high FoH DT firms show greater stability with a less drop (13.74%) than low FoH DT firms (17.56%). Moreover, high FoH DT firms exhibited a faster post-trough recovery whereas low FoH DT firms remained below baseline. This is consistent with our notion that firms with higher FoH DT adoption experience less severe revenue contractions and achieve more rapid rebounds following the onset of the pandemic.

Figure 4 illustrates the profitability trends by BoH DT level from the 2020Q1 to 2021Q1. Notably, high BoH DT firms experienced a steeper initial decline (14.3%) than low DT firms (7.6%). Meanwhile, high BoH DT firms rebounded more rapidly, whereas low BoH DT firms exhibited a more gradual recovery trajectory. These trends show that high BoH DT firms achieved a faster rebound than low BoH DT firms.

Figure 5 presents the stock price index trends of restaurant firms from 2019Q4 to 2020Q4. At the onset of the pandemic, high DT firms experienced a smaller decline (7.86%) than low DT firms (20.73%). High DT firms also demonstrated a faster rebound, surpassing their prepandemic levels by 2020Q3. Furthermore, high DT firms exhibited stock price trajectory that is more closely mirrored S&P 500 index, indicating closer alignment with broader market trends.

The primary objective of this study is to examine the effect of DT on restaurant resilience during the pandemic. Before the empirical analysis, a parallel trends test is performed to verify the key assumption of the DID approach that the treatment and control groups have parallel trends before the exogenous disruption. Figure 6 illustrates the trends of revenue, profitability and firm value drops across two groups (i.e. high and low DT firms). During these pre-shock periods, there were no statistically significant differences between the groups (p-value > 0.1), thus indicates the validity of the DID design.

Table 4 presents the results of a DID estimation to test the effect of DT on restaurant resilience. Panel A, B and C present empirical results on the effects of FoH DT, BoH DT and DT on firms’ ability to withstand declines in revenue, profitability and firm value, respectively. It should be noted that the key variable of interest is the interaction term between DT and a DROP_PRD, which reflects the causal impact of DT on performance drops during the pandemic (Abidi et al., 2023)

Consistent with the hypothesis (H1), the result (Panel A) revealed that FoH DT has a negative impact on sales revenue drop $(βFDTit×DROP_PRDt=−0.152, p<0.1)$⁠. This indicates that firms with high FoH DT firms have a smaller revenue decline than low FoH DT firms (Yang et al., 2021), suggesting that FoH DT facilitates operational stability by exploiting existing resources during the disruption (Yang et al., 2021). Panel B displays the result for the effect of BoH DT on profitability drop (H3). The result shows that BoH DT has a negative impact on profitability drop $(βBDTit×DROP_PRDt=−0.114, p<0.05)$⁠. This implies that BoH DT better stabilizes profitability by exploiting existing operational resources to maintain cost efficiency during disruption (Alt, 2021). In terms of firm value (H5), the result (Panel C) revealed that DT adversely influences firm value drop $(βDTit×DROP_PRDt=−0.102, p<0.1)$⁠. This may suggest that investors perceive high DT firms are better positioned to withstand adverse condition, thus reflect in more favorable valuation (Wu et al., 2022).

Table 5 illustrates the results of estimating the impact of DT on performance recovery in the postpandemic period. Panel D, E and F present empirical results on the effects of FoH DT, BoH DT and DT on revenue, profitability and firm value rebounds, respectively. Consistent with the hypothesis (H2), the result (Panel D) showed that FoH DT positively influences revenue rebound $(β=0.227, p<0.1)$⁠, indicating that FoH DT played a crucial role in accelerating revenue recovery following the initial shock. This suggests that FoH DT facilitates flexibility by exploring new approach that allow firms to adapt offerings in response to evolving market situation (Bilgihan and Ricci, 2024). Contrary to our hypothesis (H4), the result (Panel E) showed that BoH DT does not significantly influence profitability rebound $(β=0.053, p>0.1)$⁠. This study suspects that insignificant result may be due to a temporal misalignment between adoption and materialization of digital investment. While BoH DT enhances flexibility, the benefit may require a longer time to materialize (Stylos et al., 2021; Nam et al., 2021). Similarly, it is further shown that DT does not have a significant effect on firm value rebound $β=-0.053, p>0.1$⁠. This suggests that investors may not interpret DT as an immediate signal for recovery, partly due to asymmetric information.

To verify the robustness of our main findings, the models are reestimated using an accounting-based proxy for DT (Jiang et al., 2022). Specifically, we used the ratio of digital related intangible assets to total intangible assets (Jiang et al., 2022). It is important to note that the reestimated results are compared with the DT-firm value resilience results (Panels C and F) because this proxy pertains only to the overall DT level. As presented in Table 6, the interaction term between the DT proxy and firm value drop is negative and statistically significant $(βDTit×DROP_PRDt= -0.121, p<0.1)$⁠, consistent in both direction and significance with the prior Panels C result in Table 4. In addition, the direction of the relationship between DT proxy and firm value rebound $(βDTit= -0.116, p>0.1)$ is also aligned with Panel F result reported in Table 5. The consistent findings reinforce the robustness of our text-mining-based DT measure.

This study further examines recovery duration as a complement indicator of flexibility. Figure 7 illustrates the cumulative percentage of firms bounced back to pre-drop revenue levels. The result shows that high FoH DT firms recovered more rapidly than low DT firms: all high FoH DT firms had fully recovered their revenue to prepandemic level by 2021Q4, while only about 60% of low FoH DT firms recovered to their prepandemic level at the same time. This suggests that high FoH DT firms possess stronger rebound capacity and a shorter timespan to prepandemic performance.

Figure 8 presents the cumulative percentage of firms bounced back to prepandemic profitability levels. It is shown that all high BoH DT firms had fully recovered to their prepandemic margin level by 2021Q2, while it took longer period for low BoH DT firms to fully recover (2022Q3) from the pandemic. This may suggest that BoH DT facilitates a more efficient and sustained return to baseline profitability over time.

Figure 9 shows the cumulative percentage of firms bounced back to prepandemic stock price. Contrary to our expectation, the result indicated that both high and low DT firms showed a similar recovery duration. By 2021Q2, both groups had returned to baseline levels. This may suggest that while DT enhances internal operational and financial flexibility, its influence may not be immediately visible in capital markets.

This study investigates how DT enhances organizational resilience in the restaurant industry using the COVID-19 pandemic as a disruption context. Based on the ambidexterity perspective (Raisch et al., 2009), this study frames DT as a dual-function capability including both exploitative and exploratory functions. To provide a comprehensive analysis, resilience is evaluated across three financial dimensions: revenue, profitability and firm value.

Consistent with past studies (Liu et al., 2022), FoH DT is positively associated with both stability and flexibility (H1 and H2). FoH DT such as digital ordering, contactless payment and kiosks enhance revenue stability by exploiting existing resources to preserve operation (Yang et al., 2021). Simultaneously, FoH DT such as social media engagement, loyalty apps and CRM systems enables firms to develop (explore) new approach to adapt marketing and promotions in response to evolving market situations (Buhalis et al., 2023; Cheng et al., 2023).

In addition, BoH DT promotes stability by optimizing (exploiting) existing operational resources to maintain cost efficiency during disruption (H3) (Alt, 2021). BoH digital systems, such as automated kitchen systems, inventory tracking and labor management systems contribute to profitability stability by automating BoH processes and improving inventory control. Although BoH DT did not significantly enhance flexibility (H4), further analysis of recovery duration reveal that BoH DT firms return to prepandemic profitability levels more rapidly, suggesting a long-term resilience advantage.

In terms of firm value, DT enhances stability by signaling a firm’s preparedness and capacity to withstand shock (H5). Firms with higher levels of DT experience smaller declines in stock value during the disruption, indicating that investors perceive DT as a credible signal of organizational robustness under uncertainty. This finding suggests that DT functions not only as an operational capability but also as a market-facing signal that stabilizes investor expectations during disruption.

This study advances DT literature by reconceptualizing DT as an ambidextrous mechanism within the distinct functional domains within the restaurant rather than a uniform firm-level capability. While past studies primarily emphasize DT’s role in enhancing innovation and agility based on dynamic capabilities theory (Browder et al., 2024), they overlooked how DT simultaneously supports the exploitation of existing operational resources to stabilize performance during disruption. By identifying both exploitative (stability) and exploratory (flexibility) functions of DT, this study offers a more detailed theoretical account of how DT contributes to organizational resilience.

Second, this study demonstrates that DT ambidexterity is not uniformly activated across organizational domains. Past studies also overlooked how these ambidextrous mechanism function within the distinct operational domains of restaurant industry (e.g. He et al., 2023). By demonstrating that FoH and BoH DT activate ambidexterity through distinct mechanisms and achieve different financial outcomes, this study extends ambidexterity theory by showing that exploitation and exploration are functionally distributed rather than equally activated within restaurant firms.

Moreover, this study contributes to resilience literature by demonstrating that the performance implications of DT are outcome-specific rather than universal. The findings indicate that DT does not uniformly enhance all performance domains, thereby challenging widely accepted understanding that digital investment inherently promotes recovery (e.g. Ullah et al., 2025). Instead, the findings show that resilience effects of DT vary depending on DT domain (FoH vs BoH), ambidextrous mechanism (exploitation vs exploration) and performance dimension (revenue, profitability or firm value). By clarifying these relationships, the study advances DT research from generalized claims to a more detailed, contingency-based understanding of digital resilience.

This study offers actionable guidance for managers by clarifying how and where DT should be deployed to build resilience under disruption. First, the results show that FoH DT plays a critical dual role in protecting revenue flows and accelerating demand recovery. During external shocks that depress demand (e.g. pandemic restrictions and recession-driven customer downturns), managers should prioritize FoH DT as initial resilience tool. Technologies such as digital ordering, contactless payment, loyalty apps and CRM systems can be leveraged to protect immediate revenue flows and rapidly adjust promotions and menus as market conditions evolve (Kim et al., 2021). These technologies should be viewed not as convenience tools but as frontline resilience infrastructure that supports both short-term stability and adaptive market response.

Second, BoH DT should be viewed as a cost-stabilization tool rather than an immediate recovery lever. BoH DT strongly enhances stability via cost control and resource optimization (Law et al., 2025). During operational-side shocks (e.g. labor shortages and supply chain disruptions), managers should leverage BoH DT to maintain cost efficiency and operational continuity. However, managers should recognize that BoH DT alone is unlikely to generate an explorative rebound in profitability. To translate BoH DT into adaptive resilience, firms should complement technological investments with organizational learning and process integration initiatives (Stylos et al., 2021). This highlights that resilience-enhancing exploration in BoH area is organizationally mediated rather than technologically automatic.

Finally, the findings underscore the importance of strategic communication in converting DT into firm value resilience. While investors reward DT for its stabilizing role during crises (Wu et al., 2022), they do not immediately recognize its exploratory and recovery potential. This suggests that managers should proactively communicate DT initiatives as part of a consistent resilience narrative, emphasizing not only efficiency gains but also adaptive and data-driven capabilities. Positioning DT as a core element of the firm’s resilience strategy, rather than as isolated innovation projects, can reduce information asymmetry and enhance market confidence during periods of uncertainty.

Despite its contributions, this study has several limitations. The analysis is limited to publicly traded restaurant firms, which may limit generalizability to smaller or privately owned establishments. Future studies should expand the sample to include nonpublic and smaller firms to assess how DT effectiveness varies across different organizational sizes and resource environments. Moreover, 10-K disclosures may include elements of strategic signaling rather than fully reflecting what happens in day-to-day operations. Future studies could benefit from combining internal operational data such as store-level technology usage or digital transaction records. These indicators would help capture how digital tools are used in restaurants and provide a more accurate picture of DT. Finally, the study measures organizational resilience through revenue, profitability and firm value, potentially overlooking nonfinancial dimensions such as employee adaptability, customer retention and stakeholder trust. Future study should expand the measurement framework to incorporate qualitative and nonfinancial dimensions to capture more holistic view of resilience.