Introduction In the digital intelligence era, makers increasingly work in virtual communities, yet it remains unclear whether traditional job embeddedness constructs apply to these new contexts. This study explores the nature of virtual community job embeddedness (VCJE) and its effect on makers’ continuance participation. Methods Using a mixed-method design, we first applied grounded theory to interview and archival data from four leading Chinese maker communities to inductively develop the VCJE construct. We then followed a rigorous scale development procedure to produce a measurement instrument. Finally, we tested its predictive validity using survey data from 1,267 digital makers, employing hierarchical regression to examine VCJE’s impact on continuance participation intention while controlling for individual and environmental factors. Results VCJE emerged as a three-dimensional construct comprising Community Benefits,Job Fit, and Emotional Links, each dimension differing meaningfully from traditional job embeddedness. The final 14-item scale demonstrated strong reliability and validity. Hierarchical regression revealed that VCJE significantly and positively predicts continuance participation intention, even after accounting for established predictors such as community incentives and collaboration. Among the three dimensions, Job Fit exerted the strongest influence. Discussion This study extends job embeddedness theory into the virtual work context by redefining its core dimensions, shifting the focus from an inertial force preventing departure to a motivational force attracting continued participation. The validated VCJE scale provides a robust tool for future research on digital organizing and online labor platforms. Practically, platform operators can foster maker retention by enhancing job fit, emotional links, and community benefits. Viewed through conservation of resources and network embeddedness theories, VCJE offers a novel framework for understanding digital talent management in the Work 4.0 era.
Bi et al. (2026) studied this question.