From INFORUM 2007. Full text available. Lots of useful criticisms of Google Scholar.
Since the 1970s there has been significant competition among the major players within the traditional, subscription-based professional and academic online information services arena. Since the late 1990s some of the mushrooming (partially) open access information services which originally targeted the huge market of casual Internet users also entered the arena and posed a new challenge for the rival incumbents. The increasingly fierce competition enhances the academic and professional information services with increasingly better features. Traditional information retrieval techniques, such as citation-based searching, multiple database searching, ranking of result sets by multiple data elements, are being adopted and chiseled by the developers of the open access services. They create databases using autonomous citation indexing technology, metasearch tools for federated searching, sophisticated clustering algorithms for grouping results into topical and other subsets, link resolvers for guiding the users to complete documents, and various tools for visualizing information. Many also license content from the traditional services. In turn, these refined retrieval tools and resource enhancements keep showing up in the fee-based, professional and academic online information services, deployed in metadata-rich, highly structured mega-databases which have served academic, public, school and special libraries for decades. There are novel academic and professional information retrieval services, mashing up the best of both worlds, interlinked through one click or few clicks to deliver the most pertinent primary documents with informative metadata for the patrons. We, the end-users, will all be the beneficiaries of this rivalry -or rather, "co-opetition" - , enjoying the much improving solutions for resource discovery and result refinement, increasing recall and precision at our whim, reaping the benefits from the synergy of on-the-fly mash-ups of data from different sources in the process of searching - to turn it into finding and instant gratification.