Although the study reported in this paper is theoretically grounded and the results are based on a matched survey sample of representative firms, it has several limitations that should be acknowledged. First, the data for the current research were obtained from manufacturing firms in Northern China, raising the possibility that the results may not be directly applicable to certain Western countries. Despite this caveat, with the exception of the variable ‘‘ownership structure,’’ which may be unique to China, there appears to be no reason why the model could not be applicable to other contexts. Furthermore, our model could be relevant to firms in other emerging markets (e.g., Brazil and India). Moreover, as more non-Chinese firms shift some of their operations to China or collaborate with Chinese business partners, the findings of this study could be of importance to many firms irrespective of the physical locations of the headquarters or the exchanges on which they are listed. Second, our data sources were drawn from manufacturing firms and the possibility of contextual differences suggests that the role of IT capabilities in supporting CE activities and product innovation should be examined in other industries. Third, in this study, we took a matched survey sample approach by asking single key informants (per occupational position) to provide data for the main constructs in our model. According to established guidelines, our informants were deemed to be capable of providing useful responses. Nevertheless, future studies could consider using multi-informant designs to re-examine our model. Fourth, we adopted existing scales to measure IT capabilities as a formative second-order construct with four first-order factors. Although these four dimensions of IT capabilities were previously used successfully and are likely to explain variability in CE activities (e.g., [39,57]), future research may extend the portfolio of IT capabilities covered in this study by incorporating other IT capabilities. Finally, although our study collected data about product innovation performance one year after the collection of the antecedent constructs, we cannot conclude causal relationships with absolute certainty. More evidence based on longitudinal research is needed to further determine causation.