Knowledge-Based Dynamic Capabilities Link Digital Networks to Firm Innovation Resilience

In an era defined by volatility, uncertainty, complexity, and ambiguity—conditions that researchers now routinely abbreviate as VUCA—the ability of companies to keep innovating under stress has become one of the most consequential questions in economics and management science. A new study published in the journal Discover Sustainability offers a data-driven answer: firms that embed themselves deeply in digital innovation networks are significantly more resilient innovators, and the secret ingredient is a well-developed capacity to acquire, integrate, and reconfigure knowledge.

The research, conducted by Hu Yue and Yiqi Luo of the School of Economics at Guangdong University of Technology in Guangzhou, China, together with Jing Zhou of the Business School at Xiangtan University, constructs a theoretical framework linking network embedding, knowledge-based dynamic capability, and innovation resilience. Drawing on cooperative digital invention patent data to map digital innovation networks, and combining that mapping with panel data covering Chinese A-share listed firms from 2012 to 2023, the team tested whether a firm’s position and relationships within these networks translate into a measurable capacity to withstand and recover from shocks to its innovation output.

The core concept at the heart of the study is “innovation resilience”—the ability of a firm to maintain, absorb, and rebound its innovative activity when confronted with adverse conditions. Rather than treating resilience as an abstract quality, the researchers operationalized it through firms’ innovation performance over time, allowing them to observe which companies bent under pressure and which recovered quickly. On the other side of the equation, they distinguished between two forms of network embedding. Structural embedding refers to a firm’s overall position in the architecture of the network—how central it is, how many connections it maintains, and how it sits within the web of collaborative ties. Relational embedding, by contrast, concerns the quality of individual connections: the strength, trust, and frequency of interaction with specific partners.

The headline finding is unambiguous. Both structural and relational embedding are significantly and positively associated with firm innovation resilience. In other words, companies that occupy advantageous positions within digital innovation networks, and that maintain strong collaborative relationships within those networks, are demonstrably better at sustaining their innovative performance through turbulent periods. This holds true across a twelve-year panel spanning one of the most turbulent stretches in recent economic history, including pandemic-era disruptions and significant macroeconomic swings.

But the study goes further, probing the mechanism behind this association. The researchers hypothesized that the benefit of network embedding flows through what they call knowledge-based dynamic capability—the capacity of a firm to acquire new knowledge from its environment, integrate that knowledge with its existing internal base, and reconfigure its resources and routines accordingly. This tripartite capability, drawn from the dynamic capabilities tradition in strategic management, treats knowledge not as a static asset but as a resource that must be continuously sensed, absorbed, and recombined.

Using mediation analysis with bootstrapped confidence intervals—a statistical technique that repeatedly resamples the data to estimate the precision of indirect effects—the team confirmed that knowledge acquisition, knowledge integration, and knowledge reconfiguration each mediate the relationship between network embedding and innovation resilience. The interpretation is intuitive: a central position in a digital innovation network gives a firm access to a richer stream of external knowledge; strong relationships make that knowledge trustworthy and easier to absorb; and once acquired, the knowledge only becomes resilience if the firm can integrate it into its operations and reconfigure itself around it. The network provides the raw material, but the firm’s dynamic capability is the machinery that turns it into staying power.

Perhaps the most practically relevant results concern who benefits most. The researchers found that the resilience-enhancing effect of network embedding is stronger for three groups: firms whose executives are digitally savvy, non-high-tech firms, and—strikingly—firms operating during economic downturns. The first result suggests that leadership matters: when top management understands digital technologies, the firm is better positioned to exploit the knowledge flowing through its networks. The second is a corrective to the assumption that digital network benefits accrue only to technology-sector natives; traditional firms may actually have more to gain, since network embedding compensates for weaker internal digital knowledge bases. The third result carries an implicit policy message: in bad economic times, connections to digital innovation networks may function as a kind of insurance policy for corporate innovation, buffering firms precisely when internal resources are stretched thin. The authors verified these group differences using Chow tests, a statistical procedure for confirming that relationships differ significantly across subgroups rather than merely appearing to.

A persistent headache in this literature is endogeneity—the possibility that resilient firms simply become better connected, rather than connectivity breeding resilience, or that unobserved firm characteristics drive both. The research team addressed this with two complementary strategies. First, they employed instrumental variables, using external factors that influence a firm’s network position but have no direct bearing on its innovation resilience, to isolate the causal direction. Second, they applied Heckman two-stage models, a method designed to correct for selection bias arising from the fact that firms choose whether to participate in innovation networks at all. Both sets of checks supported the robustness of the central findings, lending considerable weight to the conclusion that network embedding genuinely causes greater innovation resilience rather than merely accompanying it.

The data infrastructure behind the study is itself notable. By extracting cooperative digital invention patent data, the researchers were able to construct actual networks of collaboration among firms around digital technologies, rather than relying on surveys or self-reported partnership data. Linking this network structure to the financial and innovation records of Chinese A-share listed firms gave the analysis both scale and temporal depth, covering 2012 to 2023—a period during which China’s digital economy expanded dramatically and firms’ innovation strategies were repeatedly tested by external shocks.

The theoretical contribution lies in bridging two research streams that have often run in parallel. Network theory has long argued that position matters: firms bridging structural holes or sitting at the center of dense clusters gain information advantages and control benefits. The dynamic capabilities framework, meanwhile, has emphasized that firms need sensing, seizing, and reconfiguring capacities to thrive in changing environments. By demonstrating that network embedding enhances resilience specifically through knowledge-based dynamic capability, the study provides the connective tissue between these traditions, showing that network position is valuable only insofar as firms possess the capabilities to exploit the knowledge it provides.

The implications reach beyond academia. For managers, the message is that cultivating digital innovation networks should be treated as a strategic resilience investment, not merely a channel for short-term technology sourcing. Building relationships with partners in digital domains, investing in the absorptive capacity to integrate external knowledge, and maintaining organizational flexibility to reconfigure resources are all concrete levers. For firms outside the high-tech sector, the findings are particularly encouraging: digital networks are not the exclusive province of technology champions, and embedding in them may disproportionately reward latecomers and traditional industries seeking to modernize their innovation processes.

For policymakers, the countercyclical pattern in the results suggests that supporting collaborative digital innovation infrastructure during economic downturns could yield outsized returns, helping firms preserve innovative capacity precisely when market failures and credit constraints would otherwise cause it to atrophy. Innovation lost during a downturn may be permanently foregone; networks appear to help firms avoid that loss.

The study, published open access and citable under a permanent DOI as part of Springer Nature’s accelerated publication pathway, was supported by the National Social Science Fund of China (grant 23&ZD090). The authors declare no competing interests. While the analysis is grounded in the Chinese listed-firm context—a setting of unusual digital dynamism—the theoretical framework of embedding, capability, and resilience is portable, and the findings are likely to prompt replication efforts in other economies. As organizations worldwide grapple with an environment that seems permanently volatile, the study offers a measured but pointed conclusion: in the digital age, innovation resilience is not built in isolation. It is built in networks, through relationships, and by firms that know how to turn borrowed knowledge into enduring adaptive strength.