It is known that end users of products and services sometimes innovate, and that innovations developed by users sometimes become the basis for important newcommercial products and services. It has also been argued and to some extent shown that such innovations will be found concentrated in a "lead user" segment of the user community. However, neither the characteristics of innovating users nor the scope of the community that they "lead" has been explored in depth. In this paper, we explore the characteristics of innovation, innovators, and innovation sharing by library users of OPAC information search systems in Australia. This market has capable users, but it is nonetheless clearly a "follower" with respect to worldwide technological advance. Wefind that 26% of users in this local market nonetheless do modify their OPACs in both major and minor ways, and that OPAC manufacturers judge many of these user modifications to be of commercial interest. We find that we can distinguish modifying from nonmodifying users on the basis of a number of factors, including their "leading-edge status" and their in-house technical capabilities. We find that many innovating users freely share their innovations with others, and find that we can distinguish users that share information about their modifications from users that do not. We conclude by considering some implications of our findings for idea generation practices in marketing.lead users, idea generation, new product development

The authors develop a model of consideration set composition. The approach taken is to compare the marginal expected benefits of including an additional brand in the consideration set with its associated costs of consideration. From an expression of the utility that a brand needs to gain membership in an existing consideration set, the authors derive an expression for set composition and optimal set size. They develop a measurement method to test the model at the individual level and apply it to the ready-to-eat cereal market. The model is tested in two ways. First, the utility function is calibrated at the individual level and the model is used to predict consideration of existing brands. The calibrated model also is used to forecast individual consideration of three new product concepts. Second, the predictive ability of a two-stage model of consideration and choice is tested against a traditional one-stage choice model. The authors conclude with a discussion of management implications of the model in terms of auditing currently available brands and new product management.

Stratigraphic and sedimentological data from New South Wales and Queensland, eastern Australia, indicate that the Late Palaeozoic Ice Age comprised at least eight discrete glacial intervals (each 1-8 Ma in duration, here termed 'glaciations'), separated by nonglacial intervals of comparable duration. These events spanned an interval from mid-Carboniferous (c. 327 Ma) to the early Late Permian (c. 260 Ma), and illustrate a pattern of increasing climatic austerity and increasingly widespread glacial ice from initial onset until an acme in the late Early Permian, followed by an opposite trend towards the final demise of glaciation in the Late Permian. The alternating glacial-nonglacial motif suggests that the Late Palaeozoic Ice Age was considerably more dynamic than previously thought. These patterns are remarkably consistent with recent interpretations of palaeofloral change, eustatic sea-level fluctuations and CO 2 -climate-glaciation relationships for this interval of time. The detailed record of alternating glacial and nonglacial climate mode disclosed herein may facilitate more closely resolved evaluations of stratigraphic records elsewhere, notably in far-field, ice-distal, northern hemisphere successions.