Evaluation of Tourism Web Sites: A Theoretical Framework Terje Aaberge Ivar Petter Grøtte Oluf Haugen Ingjerd Skogseid Svein Ølnes Western Norway Research Institute, Sogndal, Norway {taa, ipg, oha, isk, sol}@vestforsk.no Abstract The paper outlines a theoretical framework for the discussion and identification of indicators measuring the properties of web sites. In this framework we distinguish between the web site as a medium for describing an object, the web site as an object of description and the normative judgement of a web site. The framework is the basis for a project currently carried out. It aims to establish indicators for tourism web sites and use them in an open benchmarking process to enforce the improvement of the set of independent web sites marketing a given tourist destination both individually and as a resource in the collective presentation of the destination. The case studied is the Sognefjord region in Western Norway. Keywords: quality, evaluation and assessment, human computer interaction, usability, web site 1 Introduction Presentations of tourist destinations on the web are now perceived as an important marketing tool (Al-Hasan, Jwaili, & Thomas, 2003). Most actors in the tourism industry have their own web site presenting themselves and the destination. The marketing potential of the web sites is however, rarely realised. Most web sites are lacking both in visibility and usability. They are not easily found through searches in search engines and if found their use might be such that the visitor easily gives up. Also, they often do not give a precise picture of their destination, a picture that tourists visiting the region will recognise. And more seriously, the pictures they give might be diverging. There is therefore a potential for improvement. Evaluation and benchmarking provides a strategy to achieve such an improvement. The experience from two rounds of benchmarking of the web sites of all public institutions in Norway done by a team at Western Norway Research Institute the last three years shows this. Altogether about 600 sites have been assessed twice. The evaluations have been based on two different sets of indicators; the second set being more demanding than the first. The improvement from the first to the second round was significant. It turned out that the publication of the results aroused the sense of competition among the owners