In analysing and modelling the Tuscany regional system of innovation “poles”, we first analysed the network of agents in which the poles were engaged in the three years of their start-up phase of activity. By examining systematically all available information on websites we then focused on two objectives: (i) to analyse the variety of language and content that characterize the poles in their online activities; (ii) to examine the extent poles refer to the same institutions, enterprises, organizations, projects, and among these, organizations / or activities directly related to them (such as the companies managing the poles laboratories, incubators, the adherents), what we can call their “virtual network”. This paper offers a linguistic analysis of the websites produced by innovation poles in Tuscany, in order to provide a set of variables for the evaluation of their promotion strategies and to detect which are the agents involved. The focus of the analysis is the evaluation of original vs. reported information, which is here used as indicator of the innovation poles’ engagement in the promotion of their activities. The purpose is to quantify the level of engagement in their networks that each single pole has built – on its website – through the publication of texts. Through Corpus Linguistics methodology, and using both a quantitative and a qualitative approach (similar to the one proposed by the CADS framework), this papers looks at how a set of innovation intermediaries supported by a regional innovation policy have used a set of 24 terms related to the notions of innovation and promotion their core activities (business, businesses, centers, collaboration, knowledge, finance, management, manager, business, business, industrial, innovation, pole, poles, processes, projects, design, research, service, services, development, technology , technologies, territory). By looking at whether these terms appear in original or reported (produced by third-parties) texts, and at how their meanings and connotations are constructed in the former cases, this paper identifies how active have the innovation poles been in promoting their work. In addition to linguistic analysis, text analysis of web pages also allowed to identify main domains (urls) and complete links mentioned on the websites of poles. Adopting a perspective of social network analysis, with information on these virtual networks we analysed two issues relevant in modelling the system of innovation supported by the poles: through the citations of major domains, we highlight the extent to which the web sites have reported the poles connections with each other and with those involved in technology transfer; through the analysis of complete links present on the websites of individual poles we can identify to what extent the poles refer to the same information space. The presentation is structured as follows: section 1 presents the data, tools and methodology used in linguistic analysis; section 2 presents linguistic analysis of the 24 selected terms; in section 3 virtual networks are analysed; section 4 concludes with a comparative perspective on the “virtual” networks and the “real” multilayer networks in which the innovation poles are embedded and we discuss to which extent the two domains are close for the 12 poles.