Firms are increasingly opening up their innovation efforts to allow users to tap into the benefits they can offer, such as mobile data service (MDS) innovation on iOS and Google Android platforms. For this purpose, platforms typically provide toolkits to facilitate user participation, aiming to create an ecosystem for sustainable innovation. However, with the barriers to user innovation and attrition of existing innovators, it could be challenging for firms to attract and sustain users’ MDS innovation. With the possible benefits from user innovation, and considering the challenges faced, firms need to understand how to influence potential user innovators to take part and to encourage extant user innovators to innovate again. However, there is a lack of comprehensive research and understanding of what drives users’ intentions to innovate services and the differences in the antecedents of such intention between potential and actual user innovators. Further, although prior studies have suggested that toolkits can support user innovation, little research has theorized and empirically tested their influence. Motivated thus, this study proposes a model based on (1) user innovation theory to explain the antecedents (including toolkit support) of user MDS innovation intention and (2) construal level theory to explain the differential effects of the antecedents for actual and potential user innovators. We tested the model through survey data from potential and actual MDS user innovators on Google Android and iOS platforms. We find that trend leadership and anticipated extrinsic reward influence both potential and actual user innovators’ intentions to innovate. However, anticipated recognition and toolkit support affect only actual user innovators, while anticipated enjoyment affects only potential user innovators. Interestingly, toolkit support strengthens the influence of anticipated enjoyment for actual user innovators but weakens its influence for potential user innovators. Further, potential user innovators value anticipated extrinsic rewards less than actual innovators do. The implications for research and practice are discussed.