R ESEARCH INTO LIBRARIES ’ PURCHASING AND ACCESS REQUIREMENTS

Subscription agents have been involved in the distribution chain for academic periodicals for a great many years, and their contribution was reviewed recently by John Merriman and Rollo Turner. The move from print to electronic publication of scholarly journals has been swift, but there has been time for it to be documented by prominent scholars of the publishing process, such as Christine Borgman and Carol Tenopir and Donald King. However, these authors have had little to say about effects of electronic scholarly publishing on the intermediaries. For information about the current state of the subscription agency trade, one needs to use the trade literature such as Scholarly Communications Report, which has documented the rapid changes in the subscription agency business over the period for which it has been published (1997 to date), especially the mergers and takeovers that have been occurring among traditional agents, and the entry into the scholarly periodicals business of new kinds of intermediary. In these rapidly changing times, the trade association of the subscription agents and intermediaries, The Association of Subscription Agents and Intermediaries (ASA) decided to fund a study of the views of their customers (serials acquisitions librarians) about the services currently offered by agents, covering both the traditional services for acquisition of print journals and newer services concerned with electronic journals. This work was contracted to Loughborough University, and this paper reports the main findings of the survey.


Introduction
Subscription agents have been involved in the distribution chain for academic periodicals for a great many years, and their contribution was reviewed recently by John Merriman and Rollo Turner 1 .
The move from print to electronic publication of scholarly journals has been swift, but there has been time for it to be documented by prominent scholars of the publishing process, such as Christine Borgman 2 and Carol Tenopir and Donald King 3 . However, these authors have had little to say about effects of electronic scholarly publishing on the intermediaries. For information about the current state of the subscription agency trade, one needs to use the trade literature such as Scholarly Communications Report 4 , which has documented the rapid changes in the subscription agency business over the period for which it has been published (1997 to date), especially the mergers and takeovers that have been occurring among traditional agents, and the entry into the scholarly periodicals business of new kinds of intermediary.
In these rapidly changing times, the trade association of the subscription agents and intermediaries, The Association of Subscription Agents and Intermediaries (ASA) decided to fund a study of the views of their customers (serials acquisitions librarians) about the services currently offered by agents, covering both the traditional services for acquisition of print journals and newer services concerned with electronic journals. This work was contracted to Loughborough University, and this paper reports the main findings of the survey.

Aims and objectives
This research was aimed at determining what aspects of services provided by agents are valued by libraries, and how they perceive A survey of serials acquisitions librarians to ascertain their experience of and views about subscription agents and their services has been undertaken. While the services of agents with print subscriptions remain highly valued, their services are less used for electronic purchases. Although respondents would have liked to use agents for electronic journals, the financial advantages of consortium purchasing are compelling. It appears that agents should concentrate their marketing efforts on special libraries, on the North American market, and on the service of negotiating on behalf of consortia.
the role of the agent and intermediary in the purchase of and access to their subscriptions both paper and electronic. Specifically the aims of the research were to provide information in the following areas: 1. To determine how libraries organise the purchasing of their journals and their preferences for the different means of doing so.
2. To examine whether there are significant differences in these preferences between the acquisition of paper products and electronic journals.
3. To examine their likely future preferences for intermediary services in the foreseeable future, i.e. when the large majority of journals are available in electronic form.

4.
To determine those areas of the purchasing and accessing of electronic journals subscriptions which could be improved and especially those areas in which current services are considered weak.

Areas covered
The ASA specified the areas that were to be covered in the research. These included both current functions of agents in the print field, and existing and possible future roles in the electronic environment, as follows.
Purchasing: The methods currently used by libraries to purchase subscriptions and how this is changing. This section included library practices and perceptions of value.
Subscription management: How libraries currently manage their subscriptions and report to their users on them, and the role played by agents in this process.
Electronic publishing: The methods used by libraries to purchase electronic journals and to access their electronic subscriptions.
Licensing: Libraries' methods of handling many different licences, and their perceptions of good and bad practice in this area. This aspect specifically included the libraries' perceptions of agents and intermediaries in the management of licences and including consortia negotiations.

Methodology
The ASA specifically required that both academic and corporate libraries should be included, that the person responsible for serials acquisitions should be contacted at each library, and that the sample surveyed should represent an international spread of libraries.
The recruitment of this sample proved to be the most challenging aspect of the work. Because information on all of the topics mentioned above was sought, the survey instrument was necessarily long, and it proved difficult to persuade busy librarians to complete it.
To find librarians with the range of workplaces and international scope required, both personal contacts and a range of e-mail discussion lists were used. Nine individual librarians in different countries were approached and asked to distribute the request to participate to lists in their part of the world. The United Kingdom Serials Group (UKSG) also allowed us to use their list, messages were sent to several other relevant lists, and Lyndsay Rees-Jones, Workplace and Solo Adviser at the Chartered Institute of Library and Information Professionals (CILIP), provided valuable contacts to the corporate library sector. In all, requests went to over 1000 libraries.
The sample was not random, since it had been recruited purposefully to obtain a spread of countries and of library types, and the relatively low response rate implied an element of selfselection by those with a greater interest in the topic of agents' services.
The methods used to collect evidence were a questionnaire survey and a series of structured interviews conducted by telephone. The nature of the sample was taken into account in the choice of statistical methods used for analysis of the quantitative data. The Statistical Package for the Social Sciences (SPSS) was used to handle the data and produce descriptive results. For testing for statistical significance, a two-tailed t-test was generally used.
Twenty of the respondents who had agreed to take part in the study also agreed to take part in an interview by telephone. In addition, ten more interviews with UK libraries were obtained by 'cold calling'.

Sample
A total of 96 libraries responded to the questionnaire. Overall, 43% were from the UK, 29% from North America, and 26% from the Rest of Europe. Over two-thirds were academic libraries, with the rest consisting of a mixture of corporate, research institute, public and government libraries (with a further 5% defining themselves as 'other'). Thirty telephone interviews also took place.

Quantitative questionnaire results
Over 80% of respondents currently purchase their print journals from a subscription agent, but the most popular method of purchasing electronic journals was via consortia, followed by 'direct from the publisher' and then subscription agents. When asked about their preferred method of purchasing, the responses were similar, with over 80% preferring agents for printed purchases and consortia being the most popular for electronic, with agents being second.
With regard to the number of subscription agents used, just over a third of respondents used one agent and very slightly fewer used three or more agents. Over 60% of respondents had been using their main agent for more than five years, and almost three-quarters of those who responded said they received a discount from their agent for early payment. Only about one-third of respondents were able to specify the amount of the discount, and from this group, the mean discount was 2.2% (many respondents said that the discount depended on a number of factors, such as exactly how early they paid, and so on).
The vast majority of those who responded (about 80%) said that they are likely to be using the same main agent in 12 months time, and only 3% said they would not be using the same agent.
With regard to tendering, there was no overwhelming pattern in the frequency of retendering for subscription agency services. Approximately one-fifth of respondents did not answer this question and a quarter of those who did answer were unable to say how often they tender. Of the remainder, almost one-third said they re-tendered every three to four years, 16% said it was done annually, 13% said every five years or more, 9% said they never put requirements out to tender and 7% said every one to two years. Over half of respondents were not obliged to put their subscription requirements out to tender -approximately one-third said they were obliged.
When asked about the most important factors when making decisions about tendering, over 50% of those who responded said quality of service was the most important factor and just over 40% said price was the second most important factor. The size of the organisation was cited by 80% of respondents as the least important factor.
Just over a quarter of respondents said they advertised their tendering requirements by direct mailing and marginally more said they used 'other ' methods to advertise. Two or more methods of advertising were used by 13% of respondents. Over half of those who expressed an opinion believed that tendering was beneficial for their library and just over one-quarter were unsure.
Over 80% of respondents said they did not purchase any print journals via consortia, but almost two-thirds said they used consortia for purchasing electronic journals. For the small number who did purchase via consortia for printed journals, the average number of journals they purchased via this method was 1554. The number was very similar for electronic consortia purchases, at 1595 journals.

Services from subscription agents
Printed products were by far the most popular format of media purchased from subscription agents, with 25% of responses stating that they purchased six different products in printed format. For other formats, one-fifth of respondents bought two Internet products, 40% bought one CD-ROM product, 14% bought one microform product and only 3% bought one floppy disc product from subscription agents.
With regard to the products themselves, over two-thirds of respondents bought magazines in either one or two formats, one-fifth bought both journals and abstracts in three formats, almost half bought newspapers in one or two formats, with a similar figure for newsletters, and almost one-quarter bought reports in only one format.
Ratings for agents' additional services showed that 'incorporating new technology' was positively rated by over two-thirds of respondents, closely followed by 'setting up esubscriptions'. Just over half gave positive ratings for both 'supplying sample copies of print journals' and 'consolidation services'. Slightly less than half of respondents gave positive ratings for 'interface with library management systems', although it should be noted that this service did not apply to 39% of respondents.
Missing issues was rated as the most common reason for making a claim for print journals by half of respondents, closely followed by 'late issues'. 'Renewal failure' was also rated fairly highly. For electronic journals, the most common claim reason was 'access denied' (cited by almost three-quarters of respondents). Other important claim issues for electronic journals were 'technical problems with access' and 'new subscription not started'.
For over half of respondents, it took one to two months to resolve the 'average' claim, and for a quarter, it took less than one month. Less than 20% said that claims took three months or more to resolve. With regard to the acceptability of this claim time, only 40% of respondents regarded it as satisfactory, although approximately one-quarter did not actually respond to this question. However, when asked if they thought dealing directly with the publisher would result in improvements, only 21% though it would, with 37% thinking things would be worse and 41% saying they didn't know.
Almost three-quarters of those who responded gave their agent a rating of either 'excellent' or 'satisfactory' for their ability to deal with claims in general, as well as their web site. An even higher percentage (82%) gave these same positive ratings for agents' renewal lists, with the figure jumping to 90% giving positive ratings for agents' invoicing accuracy and their ability to deal with queries. There were slightly smaller percentages who gave positive ratings for agents' management reports, agents' provision of information and their ability to develop new technology, although it was still the majority opinion (61%, 66% and 66% respectively). Ratings for agents' customer service were also very positive by 90% of respondents, and their ability to deal with cancellations was rated only slightly lower, with 85% rating this as either 'excellent' or 'satisfactory'.
Three-quarters of respondents were able to specify which subscription agent service was most valuable to them, and 40% of these specified consolidation services, with a further 17% stating liaison with publishers. With regard to the importance of various aspects of agents' customer service, the following percentage of respondents gave ratings of between 'very important' and 'moderately important' to: the importance of having a named contact at the agent (99%); the agent knowing you (91%); dealing with the same person at the agent (97%); and the agent knowing the business (100%).
Ratings for agents' invoicing procedures showed more variation. Just over two-thirds of respondents rated agents' price estimations as either good or excellent, but just over one-third gave the same ratings to provision of machinereadable invoices. Approximately 90% of respondents gave positive ratings for both detailed invoices and the timing of invoices. Almost three-quarters found dealing with invoice queries to be either 'good' or 'excellent', and just over half gave positive ratings to the clarity of currency conversions. Single line invoicing was given positive ratings by over 80% of respondents. The majority of respondents (almost two-thirds) stated a preference for one main invoice, as opposed to several smaller invoices.
Half of respondents strongly agreed that agents provide a valuable service, with a further 36% agreeing with this statement. Over a third of respondents were neutral about agents' charges being easy to understand and almost half gave a positive response. With regard to consortia purchasing, 80% believed that this provides better value for money than purchasing as an individual library, and only 11% felt that they got a better quality of service when purchasing individually. Almost three-quarters of respondents either agreed or strongly agreed that setting up electronic subscriptions is a frustrating process, although just under half believed they represented good value for money (46% were neutral about value for money). Almost 80% of respondents agreed that it was easier to purchase print subscriptions via an agent, as opposed to direct from the publisher, but only a quarter agreed that the same was true of electronic subscriptions.
A third of respondents believed that aggregators were more useful than going directly to the publishers of abstract and indexing services, with over half not expressing an opinion.

Electronic journals
Over two-thirds of respondents felt that, from an administrative point of view, electronic journal subscriptions are harder to set up than print subscriptions -14% felt there was no difference. Almost half of respondents did not specify the number of electronic licences they had purchased from various sources, but from those who did respond, the mean number of licences purchased from publishers directly was 116, with 442 from consortia, 65 via independent negotiation and 70 from other third parties.
Respondents were quite evenly divided when it came to sales taxes (such as VAT) being paid on electronic subscriptions, with 42% saying they had to pay tax and 46% saying they did not -12% did not know. However, of those who pay sales tax, over half said that this did not affect their electronic purchasing decisions.
Almost half of those who responded said that ideally, they would prefer to purchase their electronic subscriptions via a subscription agent. The most popular method used to monitor use of electronic collections was publishers' statistics, with almost two-thirds of those who responded using this either frequently or regularly. This was followed by aggregators' statistics, then in-house statistics, then anecdotal information.
Loss of access to an electronic journal at the end of a subscription period, despite it having been renewed, had been a problem for over 70% of respondents. Of these, almost one-third said it had been a major problem and only 2% classified it as being 'minimal inconvenience'. Almost threequarters of respondents felt it would be helpful if publishers were to allow a 'grace' period for the first one or two issues of electronic subscriptions.
Almost half of respondents said their subscription agent had no involvement in the management of their electronic licences, and only 4% said their agent had complete responsibility for this. Almost two-thirds of those who expressed an opinion said their requirements for agents' services were likely to change as electronic journals became more readily available.
Opinions about the ease of purchasing electronic subscriptions were evenly divided, with approximately one-third being neutral, onethird in agreement and one-third disagreeing that they were easy to deal with.
However, over two-thirds of respondents agreed or strongly agreed that there are often difficulties in accessing electronic journals. The majority of respondents were neutral about whether search options were user-friendly. Just over half felt that it can be difficult to grant access to appropriate users. Almost half of respondents disagreed that agents have a high level of expertise in electronic journals. Most respondents also disagreed with the statement that there are rarely problems with service availability. Almost half of respondents felt that navigation of electronic journals can be confusing. Over 60% did not feel that licence negotiations were generally easy. Just under half of respondents felt it was not easy to find out what the licence entitles you to, and 44% felt there was a lack of compatibility between electronic journals, with a similar percentage being neutral on this. Half of respondents felt there were too many passwords to remember for electronic journals. Half of respondents did not express an opinion about whether access via a third party is easy. Almost half felt that restrictions on the number of simultaneous users were problematic.

Journals collections now and in the future
Over the next five years, respondents anticipated that their journals collections would change in the following ways. Almost two-thirds said journal holdings would reduce; almost half said newspaper holdings would reduce; 57% said newsletters would stay the same; 58% said reports would stay the same; almost threequarters said printed holdings would reduce; 96% said Internet holdings would increase; 57% said CD-ROM holdings would reduce; half said microform holdings would reduce; 60% said floppy disk holdings would stay the same; just under half said magazines would reduce.
So far as agents are concerned, 54% said there would be no change in their dealings with agents over the next two years, but 44% said they would rely on agents more over the next five years. Overall, 53% said there would be no change in direct purchasing from publishers over the next two years. But over the next five years, the question whether there would be no change in, more reliance on, or less reliance on direct purchasing produced answers of approximately one-third for each response. Almost 60% said there would be no change in their use of agents' consolidation services over the next two years, but over the next five years 46% said there would be no change and 42% said there would be more reliance on this. As regards aggregation services, half said there would be increased reliance on these and half said there would be no change over the next two years, but over the next five years there was a stronger move towards more reliance on these.
Over 40% of respondents did not know where their journals came from, but a quarter said the majority were from North America and one-fifth said the majority were from the UK.

Statistical significance
Many correlations between answers to different questions were tested using Student's t-test. A number of significant correlations were identified, but unfortunately most of these were rather unsurprising -for example, that libraries with larger numbers of users had larger budgets, and purchased more titles.

Qualitative results from the interviews and the questionnaire
The interviews were recorded and transcripts prepared. The answers to each of the questions in the interview protocol were grouped together and then summarised by inspection (not using qualitative analysis software, though this was considered). Similarly the answers to those questionnaire questions that invited comment were also inspected and summarised. The qualitative data from the questionnaire and the summaries of the interview transcripts follow.

The questionnaire
Comments about tendering. Many participants could see that tendering, at least theoretically, had certain advantages, in areas such as making libraries aware of alternatives, forcing them to evaluate their services, and so on. But tendering is "becoming less beneficial" as the numbers of agents reduce. Several people mentioned the amount of time tendering processes can take, and that perhaps a good negotiation could get the same discounts without all the work of tendering.
Most valuable service from agent. The overwhelming perception of most valuable service provided by subscription agents was consolidation of subscriptions and invoicing, and generally acting as a single point of contact between libraries and the various publishers they use. One respondent did, however, mention that having a "middle man" can be a "double edged sword" because things can take longer to resolve, but this was certainly a minority view. Other valuable services mentioned were customer service, responsiveness to queries and discounted prices.
Services currently expected from aggregators. The services currently expected from aggregators mainly related to having easy access to full text resources, with minimal number of "clicks", being kept fully updated on changes to content and having comprehensive usage statistics. Many respondents also mentioned the need for licence agreements to be streamlined as well as the need for comprehensive archiving of electronic journals.

Services expected from aggregators in five years' time.
Many respondents again mentioned "enhanced usage statistics" and "more seamless links to full text". Licensing was also mentioned by several, although one respondent rather pessimistically did not envisage standardisation of licences happening "any time soon". The need for secure archiving was also highlighted.

Aggregators compared to abstracts & indexes publishers.
One respondent said that A & I services are only useful after "considerable arduous local work", whereas another said "the functionality of the A & I services is more highly developed and meets the needs of our users better". Another respondent felt that "advantages (single user interface, greater content) outweigh disadvantages (fewer add-on services, sometimes less current) on the whole".

The interviews
When purchasing printed journals, which of the services of your subscription agent do you most value? The most common answers were consolidation and "one-stop shopping" (eight respondents), single-line invoicing (seven respondents), speed of service -though this is not always achieved (seven respondents), saving of library staff time (six respondents), saving money (four respondents), and access to publishers' addresses (four respondents).
Do you think that using subscription agents is the best way to purchase electronic journals for your library? It was difficult to generalise here, but the general impression from the interview responses is that subscription agents are far less skilled in dealing with electronic journals than they are with print. Good areas for agents to concentrate on would be licence negotiations, and verifying and maintaining access to electronic journals for their clients.
What is the best way to purchase electronic journal services, then? The two most common answers to this question were consortia (five respondents) and subscription agents (five respondents); only two respondents said "package deals from publishers", though two more pointed out that the "big deals" from publishers gave access to additional titles not bought in print, something that agents cannot do. On the other hand, one said they would go to an agent "if they gave good prices and gave value-added services".
The overall impression from the answers to this question was that a lot of users would like to use agents to handle electronic subscriptions because the one-stop shop saved them a lot of work. But experience had taught them that agents are not yet sufficiently skilled in this area to provide the quick and reliable service users have become used to in agents' work with printed journals. Publishers were also criticised; when a customer buys through an agent the publisher 's system often cannot recognise them as a customer. In any event, consortium purchasing is widespread, and the benefit of the "big deal" -access to lots of extra titles for little more cost -can be irresistible.
Would you pay extra for good services? Eight respondents, said they always went for the cheapest price, either because they were in public-sector organisations that were legally obliged to, or because they were short of money. Fifteen respondents, however, gave variously qualified "yes" answers, and one said "difficult to answer".
Do you belong to any consortium for purchasing printed journals? If "yes", what are the advantages of belonging to one? If "no", would you consider consortia purchasing in the future?
There was a slight majority of "no" over "yes" answers to this question: 16 to 10. Some of those answering "no" said that they might join a consortium for print purchasing in the future, but others pointed out that the future was likely to be electronic anyway. It was also pointed out that takeovers and mergers among agents had reduced customer choice and weakened the overall argument in favour of agents.
Do you belong to any consortium for purchasing electronic journals? If "yes", what are the advantages of belonging to one? Here a clear majority belonged to a consortium, with a "yes":"no" ratio of 16 to six. Most respondents therefore buy their electronic journals through a consortium already. The major advantages were seen to be price, better licensing terms, and access to extra content not bought in print (the "big deals"). Disadvantages of consortium purchasing were seen to be a loss of independent decision-making, the time consumed in reaching consensus within a consortium, and the fact that not all publishers are willing to deal with consortia.
Compared to print journals, do electronic journals create more work for librarians? Twenty interviewees said "yes" and five "no", with three saying "about the same". Two of those saying "yes" qualified this with the comment that setting electronic journals up in the first place was time-consuming, but that once they were up and running they gave less work than print. One commented that electronic journals need the attention of more senior and highly paid staff, as well as more staff effort overall. The additional work fell into two main classes: legalistic (licence negotiation, dealing with lots of different publishers and differing licence terms) and technical (networking and software problems needing the attention of computer staff). There was a view coming through in several answers that all publishers should use standard licence terms. It was noted that some publishers refuse to deal through agents for electronic purchases, and there were suggestions that publishers are deliberately trying to bypass agents. A number of respondents felt that agents' most useful role would be in licence negotiation.
How do your library users generally feel about electronic versus printed journals -can you generalise, or does the attitude vary greatly among different groups of users? Twenty people said that electronic journals were generally preferred; only one respondent (working in a special library with a rather older clientele) said print was preferred, and eight said that it was hard to say, or that responses were mixed or 50:50. There was complete agreement that students massively preferred electronic journals. Academic staff and researchers were seen as more variable in their views, and most respondents felt that this was a simple age factor. It was also agreed that there was a diversity among disciplines. Biomedical scientists were seen as favourable to electronic and, perhaps not surprisingly, scientists in the maths/physics/computer science/electrical engineering area were also seen as enthusiasts for electronic access. One person noted that no-one asks for new subscriptions to print journals any more. One cautionary voice pointed out that electronic-only subscriptions are more expensive than hybrid ones, owing to the levying of VAT on them.
Any other comments? It is difficult to generalise about these answers, but on the whole they are sympathetic to agents' situation, while being somewhat critical of the standard of service that agents provide in the electronic arena.

CONCLUSIONS AND RECOMMENDATIONS
Subscription agents have always done a good, and well-appreciated, job for libraries in the print journals arena, and most librarians expect to continue to use them in this role. However, it is also clear that they expect the proportion of expenditure going to print to fall and that going to electronic material to rise. Hence if agents are to prosper they need to become involved in the electronic journals delivery chain. The qualitative evidence suggests that where libraries use agents for electronic purchases, they find the service less satisfactory than that provided for print. Agentslike publishers -have found it difficult to adapt to the rapid technological changes in scholarly publishing over the last few years, and it is known, from an earlier study undertaken at Loughborough for the Department of Trade and Industry 5 , that shortages of personnel with IT skills appropriate to electronic publishing work are very much part of the problem, and form a constraint on the growth of electronic publishing.
Libraries mostly use consortia for electronic journals purchasing. Librarians would very much like agents to become involved in licence negotiations with publishers and to educate the publishers about libraries' needs. Despite the increased direct contacts between libraries and publishers that have come about as a consequence of electronic publishing, it seems that many policy decisions made by publishers are taken in ignorance of the realities in libraries. Furthermore, it is clear from responses that while dealing with agents over print journals used to be a largely clerical task, senior professionally qualified librarians have to deal with electronic journal licensing issues.
Librarians would very much like to see standardisation of licences, seamless links between abstracts and indexing services and full text, better usage statistics, and a solution to the 'appropriate copy' problem. Agents should be building up their expertise so that they can help their customers to achieve these improvements. The 'one-stop shop' facility offered by both traditional subscription agents and other kinds of intermediaries is liked by librarians and their users, but since not all publishers' titles are offered by every aggregator, the ideal of one source for all titles is not achievable. Hence, to obtain best value, libraries need to enter into a variety of deals, some with publishers, some with aggregators. The publishers' 'big deals' provide electronic access to all of the publisher 's output for the price of the former print spend. Although the questionnaire did not deal with the 'big deals' very specifically, they were mentioned spontaneously in the qualitative evidence. They have drawbacks -notably that one is tied into a three-year contract with a price escalator and little scope for cancellation -but the value for money they provide is good, especially for smaller libraries, and hard-pressed librarians cannot ignore this.
North American libraries are better funded and larger than the rest, and UK libraries are worse funded and smaller than the rest. But UK libraries are more likely to use an agent for consortia purchasing, perhaps because of the existence of NESLI. As one would expect, special libraries are smaller than academic libraries.
Taking these various conclusions together, one can see that academic libraries -traditionally the agents' biggest customers -are likely to buy their electronic products through consortia and publishers' 'big deals', and since they are larger and better staffed, they are likely to have skilled senior staff able to undertake such negotiations. Special libraries, whose information needs are specialised, are unlikely to be attracted to 'big deals' which would give them access to electronic journals for which they have no conceivable need. Furthermore they are likely to be sparsely staffed, with no time to get involved in complex consortium negotiations.

Recommendations to agents
1. Agents should ensure that they recruit or retrain staff to ensure that they have the appropriate skills to deal with electronic periodicals as efficiently as they have always dealt with print ones.
2. Agents should offer their services pro-actively to library consortia, to act as the agent for a consortium in its negotiations with publishers.
3. In negotiations with publishers, agents should seek very strongly to achieve contract terms and licence conditions that are standardised and which allow libraries to provide services to their users in an acceptable way. The education of publishers about the realities in libraries is part of the intermediary function of agents.
4. Agents should market their services proactively to special and corporate libraries, which in the current climate are likely to be very lightly staffed.
5. Agents should aim their marketing efforts at the North American and continental European markets, which have greater purchasing power than the UK market and currently make less use of agents in consortium negotiations than the UK does.