As the Internet has enhanced the collection and provision of citation, usage and access metrics, the challenge lies neither in the technology nor the method, but in constructing databases that deliver services of value to the scholar. However, the development of metrics has hitherto been driven by the needs of external research assessment (governments and funders), while publishers and libraries have focused on their own needs (e.g. journal impact and usage factors). Scholars often criticize research assessment and the use of particular metrics as a zero-sum game whose undesirable consequences far outweigh the benefits. However, this is not to be confused with a general prejudice against metrics, which are principally compatible with the scholarly recognition and rewards system. But it does indicate that current metric information services often do not serve the needs of scholars. The question everybody should be asking is: What kind of metric information services would serve scholars? The argument proceeds in six steps. First, the problematic and controversial nature of assessment metrics is discussed. Second, the limited value of current metric information services is outlined. Third, the notion of metrics as research information services is clarified. Fourth, some examples of such services are offered. Fifth, the potential value is sketched from the perspective of a postdoc. Sixth, it is indicated that societies and publishers could begin building more metric information services since tried-and-tested technology and methods are available already.

Can the methods of science be directed toward science itself? How did it happen that scientists, scientific documents, and their bibliographic links came to be regarded as mathematical variables in abstract models of scientific communication? What is the role of quantitative analyses of scientific and technical documentation in current science policy and management? Bibliometrics and Citation Analysis: From the Science Citation Index to Cybermetrics answers these questions through a comprehensive overview of theories, techniques, concepts, and applications in the interdisciplinary and steadily growing field of bibliometrics.

In 2007 the United Kingdom Serials Group (UKSG), in association with the online usage metrics organisation COUNTER, published the results of a wide-ranging study which explored how online journal usage statistics might form the basis of a new metric of journal quality. The study combined a web-based survey of opinion with a series of in-depth interviews with stakeholders from the author, publisher and librarian communities. The aim of these twin avenues of research was to examine the ways in which journal quality is currently assessed, and the degree to which any additional usage-based metrics might prove valuable to each stakeholder community, along with practical ways in which such metrics might be derived and constructed to provide the maximum utility for all, within defined resource constraints. Building upon the encouraging reactions revealed in the market research, Stage 2 of the project is developing a programme of data modelling and analysis that will use real usage data from a number of content providers, with the aim of identifying potential candidate metrics for longer term scaled up testing.

Conclusion Citation counts can be reliably predicted at two years using data within three weeks of publication.