In addition, the analysis would have to take interactivity into account, not only as a matter of principle, so as to do justice to the rich affordances of the medium, but also because of the social significance that the exploitation of these affordances may have for a particular social practice. In the case of the government’s ‘benefit fraud’ campaign, for example, visitors to the website are encouraged to ‘report a cheat’ by filling in a detailed online form. The interactive potential of the medium is thus harnessed to what seems like an invidious perversion of e-government.

Furthermore, the unique semiotic potential of the web creates new challenges for the analysis of multimodality (Lemke, 2002). To return briefly to my earlier snapshots of higher education discourse, universities can now be seen to use not only colourful graphics on their websites, but also sound and movement.
(…)
The methodological challenge involved is compounded when it is not only the multimodality of web communication itself that is at issue, but its complex links with other media. This is the case with the world-wide cross-media franchises discussed by Lemke (2004), who proposes to extend CDA in the direction of a ‘multiplicative, heteroglossic model of meaning effects across media’.
Clearly, web technology has significantly extended the discursive repertoire available, allowing new modes of representation as well as the construction of new identities and relationships in ‘parallel social realities’ (Lemke, 2004). To expect these to feature prominently in every CDA project would be as unrealistic as it would be misguided. However, complex online multimodalities should at least be considered as possible areas of investigation when research questions are formulated, methods developed, and project designs drawn up.