If web-based technology is increasingly becoming the central plank of contemporary teaching and learning processes, there is still too little evidence to suggest that it is delivering purposeful learning activities beyond its widely perceived potential as a learning resource providing content and learning objects. This is due in part to the "bandwagon" effect created by the ubiquity and popularity of the web but also because e-learning is being institutionally managed and pedagogically harnessed without serious design considerations being given to its interactive specificity, which only seem to manifest themselves through sporadic and mainly inconsequential evaluation. On this premise, this paper attempts to identify some of the main design issues involved at both conceptual and implementation levels whilst making the case for the necessary collection and subsequent analysis of valid data. Therefore, it will re-examine the role, value and means of exploitation of existing evaluative data within the design process in order to facilitate the adoption of a more appropriate conceptual approach and to better understand web-based interaction in relation to learner requirements through the application of activity theory. The ulterior motive behind this study is to show the importance and relevance of the conceptual understanding of the web as a learning interactive construct and how the learning process it is meant to generate, can be improved through targeted and integrated evaluation. (Contains 2 figures.)