Increasingly, library and information services are under pressure to demonstrate value for money. Against the backdrop of search engine dominance, economic instability, and rapid technological development, Intute, a JISC-funded free national service which delivers the best of the Web for education and research, is facing reduced funding and an uncertain future. This article will share its successes and achievements, put a spotlight on the human expertise of its contributors and partners, and reflect on lessons learnt in the context of the sustainability of library and information services.
Did you use the Internet in the early 1990s? Do you remember the beginnings of the World Wide Web, the first search engines, and early Web browsers like Mosaic and Netscape Navigator? Although launched in 2006, the history of Intute takes us right back to this period; to the beginnings of the first academic subject gateways, the development of early online tutorials like the Internet Detective, and the creation of the Resource Discovery Network, (which would evolve into Intute). During the 1990s, use of the Internet grew rapidly, with increasing usage for both leisure and academic purposes, but in those early years it was difficult to find Internet-based academic content. Search engines and Web directories such as AltaVista and Yahoo had arrived on the scene to help people navigate through this new and confusing world, but there was a need for services that would bring together academic content and provide simple access for students, librarians, researchers and academics.
The initiation of the JISC Electronic Libraries (eLib) programme in 1995 led to the creation of a suite of subject gateways, and services such as SOSIG and EEVL were conceived. As the eLib funding cycle neared completion, JISC recognised that there was a need to coordinate the activities of the (now established) gateways, and the Resource Discovery Network (RDN) was formed in 1998 . The RDN ultimately comprised eight distinct subject gateways:
The Virtual Training Suite, a series of tutorials promoting Internet research skills funded under the JISC 5/99 Learning and Teaching Programme, became part of the core service in 2002.
In 2005, an Executive was formed within Mimas at The University of Manchester, and took on the challenge to make the organisation more efficient and integrated. The content of the gateways was merged into one simple interface and service, and Intute was born. The distributed organisational model became more closely integrated, drawing on the strengths and expertise of all the partners. The core of the service itself was a newly integrated database of hand-picked, carefully described Web sites. The Virtual Training Suite stood alongside this catalogue delivering its subject-based Internet research skills tutorials, and during this period, the JISC-funded Informs online tutorial creation software was transferred from the University of Huddersfield to join Intute's suite of training services.
Intute did not rest on its laurels. Over the following years, the Web site would be redeveloped and reorganised, replacing the initial set of 4 subject groupings (arts and humanities; social sciences; science, engineering and technology; health and life sciences) with a structure based around 19 top-level subject headings. This approach allowed users to focus their searching more keenly on a specific subject whilst retaining the ability to search Intute as a whole; response to these changes was extremely positive.
The context that Intute operates in has changed enormously since the early RDN days. When Google arrived after the millennium, it would quickly become the weapon of choice when searching against a backdrop of exponentially increasing content on the Web. Increasingly Google and Google Scholar offer effective search results as they index databases and reach deeper into the hidden Web, and libraries and publishers now strategically engage with them to expose their content as they cover more and more of the world's information. However, in education this is not without issues. Search engine-ranking mechanisms do not guarantee quality and do not discriminate between the biased, the commercial, and the self-interested Web site. Concerns about the student use of Internet have been explored by Tara Brabazon  and more recently by Andrew Whitworth in his book Information Obesity. Whitworth describes the scholarly environment as one where there is an increasing volume and complexity of information, and suggests that, 'finding information is no longer the problem, but being discriminating, filtering it out, and managing it is.' 
This is the problem that Intute has always responded to in its aim to facilitate the development of the critical evaluation skills required to assess the copious stream of content generated by the Web, and provide recommendations of Web sites from trusted sources.
For us our achievements have not simply been in delivering a widely used service with an evolving range of features, but also in the way we, as an organisation, have worked together and used our distributed network of expertise. In addition to the seven universities which formed the Intute consortium, more than 70 other partners were recognised at the time Intute launched. These included such prestigious bodies as the Wellcome Trust, the British Library, the Royal Geographical Society, the Royal College of Nursing, 12 of the Higher Education Academy Subject Centres and the Royal Society of Chemistry. Intute was underpinned by a wealth of influential supporters and organisations.
Creating and developing Intute has required extensive technical expertise but also an increased level of distributed teamwork, an area for which Intute has subsequently become renowned. Throughout its history, Intute partners have worked together to offer real benefits to the library and information community. We have developed machine-to-machine interfaces and worked closely with institutions to embed services and features into learning and teaching materials via the Intute Integration Project . University libraries such as Leeds have used the MyIntute personalisation interface, newsfeeds and search widgets to build online lists of resources across a range of academic subjects, and by 2009, 67% of UK universities had widely integrated Intute content into their library Web pages, VLEs or federated search systems.
Intute has continued to be technically innovative in a service environment, adding a number of Web 2.0-related enhancements, such as user comments and ratings, and the ability to filter searches in increasingly complex ways. The service also became searchable through Linden Lab's Second Life virtual world in 2007, one of the first academic services to appear there.
Our work was rewarded with greater visibility in the national media, with a Guardian headline  suggesting 'Google, eat your heart out', and pointing out that 'Intute offers what search engines don't - commitment to quality'. With a favourable review on the BBC Click  programme following, it was clear that the service was making huge strides. This attention was also reflected in the Intute usage statistics. An average month in 2008 would see more than 11 million page views, over 1 million browses and upwards of 1 million searches performed on Intute, and these figures would increase significantly in 2009. Across the academic year for 2008/9, 34.5 million keyword searches alone were carried out.
The Virtual Training Suite also went from strength to strength. Usage continued to grow apace, with almost 700,000 visitors viewing 7.6 million pages worth of tutorials in 2009. Thousands of positive endorsements were received from tutorial users: 'self explanatory, informative, easy to understand, straight to the point', 'very well organized and easy to navigate, and surprisingly interesting to follow', 'I like how explicit the tutorial was in examining the credibility and relevance of online sources'.
In 2007, Intute was awarded the Jason Farradane Award  for outstanding achievement in the information field and was described as 'a great example of the UK library community taking a long-term pioneering role in the Internet information environment, developing a national service through collaboration, which has grown to become well respected and highly used worldwide.' The achievements of service director Caroline Williams were particularly highlighted, with the committee noting her role in developing 'organisational strategies and systems that enable distributed teams to run a unified and coherent service'.
Over the years, Intute has been involved with a range of other projects, harnessing core staff expertise, technology, and using our database as a test-bed for new services and methods. At the same time, work on these cutting-edge pieces of research has enhanced Intute's core services.
Intute's research and related projects have typically built on existing services, but extended these through project funding both as demonstrators and as real enhancements to existing services. The mobile Internet Detective  is one example of a well used service extended through new technology.
Still ongoing is the ViM (Value for money in automatic Metadata generation) Project, looking for ways to maximise quality whilst minimising costs of metadata creation. Final data is not yet available, but the study has been comparing the efficiency of manual (human-edited) approaches to metadata creation, against automated tools, and looking in more detail at the kinds of metadata required by users. Initial results are surprising, suggesting that automation does not necessarily save costs – but detailed conclusions are likely to provide a legacy applicable across a broad spectrum of future data services.
Drawing on Intute's experience of integrating the varied databases and working practices of the former RDN hubs, the AIRDIP (Academy Intute Resource Database Integration) Project was a partnership with the Higher Education Academy (HEA). This integrated the online resources of ten Academy subject centres into a single database, to provide tailored subsets of data for their different user groups and configured search options to meet user needs. Against a backdrop of changing project specifications, and organisational change, this project offered a chance to learn the lesson that, according to project manager Diana Massam, 'a focus on deliverables which can be achieved and have some benefits for partners can lead to a positive outcome!' Other collaborations have included the PERTAINS (PERsonalised TAgging Information Services) Project (with the University of Glamorgan) which looked at ways of suggesting controlled and uncontrolled vocabularies for user tagging of information, and EASTER (Evaluating Automated Subject Tools for Enhancing Retrieval) (with UKOLN) to evaluate existing tools for automated subject metadata generation.
Intute's knowledge of the resource landscape in particular subject areas proved a further fertile arena for developing projects. MEDEV for Medicine, Veterinary Science and Dentistry (another collaboration with the HEA) and Organic.Edunet for agriculture (an EU-wide programme), involved building up a picture of the best and most relevant Web content in specialist fields. This subject expertise was perhaps tested most by IJDDiP (Intute/JISC Digitisation Dissemination Project) which through an Intute-organised series of academic seminars and subject-focused Web guides, aimed to generate awareness of the new JISC digitisation projects amongst the relevant sections of the UK academic research community. Here, the project recognised that 'academic researchers can be a difficult group to alert to new digitised content, as the usual dissemination channels for such content (leaflets, blogs, presentations, etc) do not always penetrate the forums of academic discussion that the researchers themselves inhabit and use to exchange information' . Although Intute's own expertise in harnessing both conventional digital and academic channels was applied, perhaps the benefits flowed mainly from the role of Intute as a mentor, facilitating best practice and offering the chance to experiment and encourage others to showcase the results.
In July 2010, we find ourselves in less celebratory times. As we enter into a one year maintenance funding phase, and begin to wind down the service in its current form, we would like to share some reflections on what went wrong for Intute.
The impact of the global economic crisis and subsequent reductions in public funding has thrust services such as ours into new and unfamiliar environments. Since the launch of Intute, the threat of funding reduction has been a constant risk, and over the last four years the ESRC, the AHRC and now primary funder JISC have withdrawn support. At the end of 2006 we commissioned an independent business development consultancy to explore alternative business models and provide advice. Their frustrating, and yet unsurprising, recommendation was that grant funding was the most appropriate fit for the core mission and values of Intute. Their market research suggested limited interest in subscription to Intute in its entirety, a point which is perhaps not surprising in the context of reducing library budgets, and proposed that sponsorship and advertising opportunities should be investigated further. Unfortunately, the economic instability that had resulted in reduced grant funding also hampered our investigations into advertising and our attempts were unsuccessful. Our exploration of alternative business models was affected by the lack of availability of business development expertise within Intute. The work of Ithaka on behalf of Strategic Content Alliance  did illustrate imaginative ways of sustaining online resources yet we struggled to translate this experience into our context, and our two attempts to recruit a business development officer failed.
Intute was successful in reducing costs so that a 15% tapering annual funding allocation was accommodated without any degradation of service, but when the news of the 80% funding cut reached us we realised we were dealing with not just an issue of efficiency but of radical transformation or demise. Our unique selling point of human selection and generation of descriptions of Web sites was a costly model, and seemed somewhat at odds with the current trend for Web 2.0 technologies and free contribution on the Internet. The way forward was not clear, but developing a community-generated model seemed like the only way to go.
So, in pursuit of sustainable routes for the Intute catalogue, we now enter into discussion about volunteers and free contribution. Clay Shirky outlines what he sees as the three different approaches to organised activity: institutional or managed action; no action (because the costs of organising outweigh the benefits of that activity); and social tools, which provide a third option: 'action by loosely structured groups, operating without managerial direction and outside the profit motive.' For Intute, if the cost of managed action outweighs the benefits, then can the third alternative way forward work for us (or indeed is it time to stop and take no action)? The success of initiatives such as Wikipedia and suggests that there is mileage in an easily editable community-driven model. Internally, although excited by the opportunities and full of innovative ideas, there was also scepticism. Intute has an existing network of potential contributors from its Twitter channels, MyIntute subscribers and the comments and rating system, but moving to this model poses fundamental questions:
We have explored the motivation to contribute and found contrasting evidence. Jakob Nielsen has well documented what he calls 'Participation Inequality' and the 90-9-1 rule , however the Intute user survey last summer painted a more optimistic picture for community contribution revealing that 54% of respondents (who were largely from the library community) would contribute their own favourite Web resources. Many respondents stated that they would need to maintain their own lists of favourite resources if they did not have the Intute IRC to rely on, and could be encouraged to contribute to a collaborative resource. The market research findings did, however, suggest that a quality-assurance process would be essential in order to maintain trust in the service and also pointed to the need for incentives to encourage contributions. If this approach was deemed to be a way forward then we would have to explore financing of low-cost incentives for contributors of accepted resources.
Referring back to our earlier question, if people do contribute, what does this mean for our current collection development processes and for our curation expertise? Clay Shirky offers the example of iStockphoto to illustrate how bringing together amateur and professional photographers can change social definitions that are not tied to professions . But where does the librarian and the expert fit in all of this? Are we grappling with new perceptions of trust and quality?
For this approach to work at Intute, our role as the expert would need to turn into one of mediator: a visible person in the network who would motivate contribution by issuing calls for action. There are compelling reasons to pursue this route, however, we are still searching for an alternative business model to support this work.
In terms of future directions for the Virtual Training Suite and Informs, the picture is clearer. Our market research indicated that development of these services to include greater customisation and enhanced support could support the introduction of a subscription model. The 2010/11 academic year will see staff involved in these services venture into new commercial territory as they explore options to keep these popular services running.
The problems that led to the creation of the RDN hubs, that is the need to find quality resources and make sense of the Internet, are still pressing issues today, and throughout its history Intute has helped students to make discerning use of the Internet through community collaboration. However, technological developments, changing user expectations and diminishing budgets mean that services such as Intute will need to find new ways to engage with their communities, and the search for alternative business models will require new ways of thinking. As Intute begins to wind down the service in its current form, we would like to thank all of the staff and partners that have worked together to develop the service throughout its history, and we hope that this tradition of collaboration will leave a legacy for the future development of community-driven resources.