Combining digital and legacy technologies: firm digital transformation strategies—evidence from Chinese manufacturing companies

The rapid digitization accelerated by the COVID-19 pandemic highlights the urgent need for effective digital transformation strategies. Digital transformation represents a pervasive trend across sectors, driving economic and technological shifts at organizational and industry levels. Successful implementation can yield adaptability, financial returns, and sustainable growth, yet many firms struggle with execution, as evidenced by a SAP survey indicating that although 84% of companies view digital transformation as critical within five years, only 3% have fully transformed. Research has explored antecedents, processes, and outcomes of digital transformation, but contradictions and gaps remain regarding how digital technologies influence existing technologies within firms. This study proposes that a firm’s digital technology strategy—defined as the deliberate application of digital technology to strengthen legacy technology—shapes transformation outcomes. We argue that sustainable digital transformation depends on how well digital technologies integrate with legacy technologies. We develop four strategies—searching, enhancing, grafting, and integrating—and examine three critical antecedents influencing strategy choice: technology distance (general vs. specific), technical search type (breadth vs. depth), and leadership perceptions (perceived usefulness and perceived ease of use). We empirically analyze 76 Chinese high-tech manufacturing enterprises using interviews and fsQCA to identify configurational effects and pathways to each strategy. The study contributes a process-oriented view of digital technology adoption, a typology of strategies, and insights into the configurative effects of distance, search, and leadership perceptions on strategic choice.

The paper situates digital transformation within literature emphasizing digital technologies’ generative, transferable, and combinable properties, which can both strengthen and reshape firms’ capabilities, products, services, and interorganizational relationships. Prior work highlights potential misalignment between digital and legacy technologies and warns that piecemeal adoption may erode advantage; instead, firms should leverage digital technologies to reinforce and recombine existing knowledge and competencies. The authors conceptualize digital technology strategy as the use of digital technology to strengthen legacy technology and develop a four-part typology: (1) Searching—finding new technologies with functions similar to legacy technology; (2) Enhancing—extending and strengthening legacy technology using digital tools; (3) Grafting—directly connecting digital and traditional technologies to add functionality; (4) Integrating—fusing digital and legacy technologies to create new combined technologies. The study theorizes configurative contexts for strategy choice based on three antecedents. Technology distance: general (replicable, transferable, widely accessible; more distant from legacy) versus specific (complex, firm-specific, harder to imitate; closer). Hypotheses H1a/H1b propose general technologies align with searching/enhancing and specific technologies with grafting/integrating. Technical search type: search breadth (external, diverse sources beyond the firm’s trajectory) and search depth (internal or closely related to the technological trajectory). H2a/H2b propose breadth aligns with searching/enhancing and depth with grafting/integrating. Leadership perceptions: perceived usefulness (performance gains) and perceived ease of use (effort required). H3a/H3b propose usefulness aligns with searching/enhancing and ease of use with grafting/integrating. Table 2 in the paper summarizes these hypothesized configurations for each strategy.

Research design: multicase qualitative study of Chinese high-tech manufacturing SMEs, complemented by fsQCA. Context: high-tech manufacturing is central to Industry 4.0 and China’s Made in China 2025; firms face urgent transformation needs and high technology investment risks. Sampling: from MIIT’s 2015–2016 list of high-tech manufacturers. Selection criteria: SMEs (<1000 employees) to capture dominant strategy choice; A-share listed firms for data availability. Final sample: 76 firms across electrical equipment, mechanical engineering, textile engineering, chemical/pharmaceuticals, and others. Data collection: 2–4 hour face-to-face interviews (2017–2018) with directors/executives on digital technology use, legacy technology development, technical search processes, and leadership perceptions; supplemented with annual reports and public documents. Coding: interviews recorded, transcribed, and systematically coded. Digital technology strategy identification used MD&A text analysis with refined deductive/inductive word lists (CATA), expert-reviewed vocabulary, and keyword counts, with examples provided. Measures (Likert 1–5): Technology distance—General technology (low acquisition cost; widely used; rapid updates), Specific technology (difficult to imitate; industry-leading; hard to transfer). Technical search type—Search breadth (varied channels; different partners; seeking new opportunities), Search depth (aligned with trajectory; from competitors). Leadership perceptions—Perceived usefulness (improves performance, productivity, effectiveness; overall usefulness), Perceived ease of use (clear/understandable; low mental effort; easy to use; easy to control). Calibration for fsQCA: average-based cutoffs; for general tech (crossover 3.2; full membership 4.1; full non-membership 1.0); specific tech (3.5; 4.6; 2.5); search breadth (3.4; 4.5; 2.2); search depth (3.6; 4.4; 2.1); perceived usefulness (3.3; 4.4; 2.3); perceived ease of use (2.9; 4.2; 2.0). Analytical approach: fuzzy set qualitative comparative analysis (fsQCA) to identify configurations linking antecedent conditions to each strategy; consistency threshold ≥0.80. Robustness: independent dual-coding plus third-party coder to mitigate bias; probit regressions run separately for each strategy as binary dependent variables to triangulate fsQCA findings.

Sample profile: 76 firms; strategy adoption—Searching 18 (23.7%), Enhancing 26 (34.2%), Grafting 18 (23.7%), Integrating 14 (18.4%). Correlations (n=76): General technology positively relates to Searching (r=0.224, p<0.05) and negatively to Integrating (r=−0.213, p<0.05). Specific technology positively relates to Grafting (r=0.252, p<0.05) and Integrating (r=0.373, p<0.01). Search breadth positively relates to Searching (r=0.461, p<0.01) and Enhancing (r=0.411, p<0.01). Search depth positively relates to Grafting (r=0.405, p<0.01) and Integrating (r=0.234, p<0.05). Perceived usefulness positively relates to Searching (r=0.261, p<0.01). Perceived ease of use positively relates to Grafting (r=0.304, p<0.01) and Integrating (r=0.387, p<0.01). fsQCA results: • Searching (3 configurations: C1a–C1c): General technology, search breadth, and perceived usefulness commonly present (often core). C1a: core—general tech, search breadth; peripheral—perceived usefulness, perceived ease of use; search depth absent. C1b: core—search breadth, perceived usefulness; peripheral—general/specific tech, perceived ease of use; search depth absent. C1c: core—general tech, perceived usefulness; perceived ease of use absent; search breadth/depth peripheral. • Enhancing (3 configurations: C2a–C2c): Search breadth central. C2a: core—search breadth, perceived ease of use; peripheral—specific tech, perceived usefulness; search depth absent. C2b: core—general tech, search breadth; peripheral—search depth, perceived ease of use. C2c: core—specific tech, perceived usefulness; search depth absent; general tech, search breadth, perceived ease of use peripheral. • Grafting (2 configurations: C3a–C3b): Search depth core in both. C3a: core—specific tech, search depth; search breadth absent; peripheral—perceived usefulness, perceived ease of use; general tech absent. C3b: core—search breadth, search depth; general tech absent; peripheral—specific tech, perceived usefulness; perceived ease of use absent. • Integrating (2 configurations: C4a–C4b): Core—search depth and perceived ease of use. C4a: core—search breadth, search depth, perceived ease of use; peripheral—specific tech, perceived usefulness. C4b: core—specific tech, search depth, perceived usefulness, perceived ease of use; peripheral—search breadth; general tech absent. Overall supports H1a/H1b (distance), H2a/H2b (search type), H3a and partial H3b (leadership perceptions). Robustness (probit): Patterns align with fsQCA—specific technology positively associated with Grafting and Integrating; search breadth with Searching and Enhancing; perceived usefulness with Searching/Enhancing; perceived ease of use positively with Integrating and negatively with Enhancing.

The findings show that firms’ digital technology strategies depend on the interplay of technology distance, search behavior, and leadership perceptions. General digital technologies, when pursued through broad external searches and perceived as useful, tend to drive Searching and Enhancing, helping firms either replace or extend legacy technologies to bolster competitiveness. Specific technologies, closer to legacy technologies, in conjunction with deep searches and leaders’ perceptions of ease of use, facilitate Grafting and Integrating, enabling close coupling or fusion with legacy systems. The configurational perspective clarifies multiple viable paths (equifinality) to the same strategic outcome, addressing uncertainty about how digital technologies interact with legacy systems. These insights inform managerial decision-making on matching digital initiatives to legacy bases and organizational search capabilities, thereby improving the likelihood of successful transformation and sustained competitive advantage.

This study contributes a process-oriented framework for digital technology adoption in transformation by: (1) positioning firms’ competitive advantage as arising from the combination of digital and legacy technologies; (2) proposing a four-strategy typology—searching, enhancing, grafting, integrating—detailing how digital technologies strengthen legacy technologies; and (3) demonstrating, via fsQCA, multiple configurations of technology distance, search types, and leadership perceptions that lead to each strategy, highlighting equifinality. Managerially, firms should assess legacy technology, align search strategies (breadth vs. depth) with technology distance (general vs. specific), and incorporate leaders’ perceptions (usefulness, ease of use) when selecting strategies. Future research should broaden contexts (industries, countries), examine additional organizational and cultural antecedents, and explore emerging digital technologies (e.g., generative AI, embodied cognition) and their roles in strategy selection and integration dynamics.

(1) External validity: the sample is limited to Chinese high-tech manufacturing SMEs; results may not generalize to other industries or countries. (2) Methodological scope: fsQCA is case-based; while suitable for configurational insights, broader generalizability requires further validation. (3) Omitted variables: only three antecedents (technology distance, search type, leadership perceptions) were examined; other organizational and cultural factors may influence strategy choice. The study also notes unexplored links to generative AI and embodied cognitive technologies as avenues for future inquiry.