Posted October 23, 2003
Libraries are spending an increasing percentage of their budgets on database subscriptions, including indexes and abstracts, collections of full-text journals and newspapers, and reference tools. It is not uncommon for a small library to spend thousands of dollars a year, and for a mid-size library to spend tens of thousands to hundreds of thousands of dollars a year. A few large libraries spend millions of dollars a year on as many as 250 databases. Given the size of the investment, libraries want to realize the best possible terms before signing agreements. As might be expected, the best possible terms are seldom realized by signing database vendors’ standard agreements.
Unfortunately, it is becoming increasingly difficult to negotiate with database vendors. A decade ago, before most vendors had standard contracts, negotiation was routine, although often limited to scope and price. One reason why negotiation is more difficult today is that database vendors have more experience. They know their costs and their market. They have tried various packagings and pricing models for their products. Also, given the number of customers already under contract, few new accounts are attractive enough to make many concessions. Most important of all is that many databases are now vended by "aggregators," rather than the database producers.
Aggregators offer the products of a number of producers, often combining them into packages. Ebsco, Gale, OCLC, Ovid, Science Direct, and SilverPlatter are some of the major aggregators. For database producers, the advantages of dealing with aggregators are that they make the investment in the hardware, software, marketing, and other costs of serving the subscribers; for libraries, the advantages of contracting with aggregators is that they handle the licensing issues with the database producers, provide the necessary patron authentication, and simplify the user interface. In many cases, aggregators even facilitated searching across multiple databases. Aggregators also serve both producers and libraries by combining complementary products in packages that are easier to sell and buy.
As aggregators gained experience in the mid-nineties, they began to resist contract negotiation and require the signing of standard contracts. All of the terms, including prices, were set forth and customers were told that there could be no negotiation. There were exceptions, however. Very large libraries and consortia were able to set aside or modify terms, including prices. As recently as 2002, a major public library was able to negotiate both a free trial period of several months and a major discount for the next two years. Consortia, especially statewide ones, have also negotiated significant changes in standard contracts.
It is still possible for smaller libraries to negotiate terms, but they must look beyond the obvious providers. Contracting directly with the database producer is one option; contracting with a small and/or relatively new aggregator is another. There is almost always an aggregator that is seeking to increase market share at the expense of an established aggregator that has become inflexible.
A library--or even a small consortium-that decides to bypass the large aggregators must be prepared for multiple negotiations in order to obtain all of the databases it wants. It must also be prepared to address the issues of patron authentication and a common user interface to a variety of products. With the advent of portal technology, two of these issues are now relatively simple except in a few cases where database producers insist on their own authentication procedures. Not only do most vendors of automated library systems offer a portal that includes a multi-protocol single search, patron authentication, and the linking to related resources; it is also possible for a library to contract directly with MuseGlobal or WebFeat for portal technology.
If the decision is to negotiate, the next step is to identify the negotiating topics. The following should be considered:
What is actually included? If the database is an online version of a print product, any difference between the print and online products should be spelled out. If it is a database of journals, the contract should include an itemized schedule of the titles, years of coverage, and format (full-text or image). It is a good idea to ascertain whether the package can be modified to exclude subject areas or specific titles that do not meet the needs of the library.
Most products are not static. It is typical for the coverage to expand each year because the amount of available information increases, but also because many products are packages of several products that are frequently modified. This is particularly true with full-text databases of journals. There should, therefore, be an effort to control "coverage creep." If it is not possible to limit the product to that which was initially negotiated, there should be a cap on the price increases that may be associated with broadened coverage. At a minimum, there should be a formula that limits the price increase to the same percentage as that which the expansion in coverage represents.
Even more important than the frequency of updates is the time lag between initial production and availability for online access. It should be spelled out. An online product should be available no later than the print product.
The library or consortium, typically called the "licensee," must continue to have access to that for which it has paid. If older content ceases to be available online, it must be available in another format. Also, if the subscription is terminated, the licensee should continue to have access to that for which it previously paid. It should, of course, expect to pay for the cost of access or the medium (CD-ROM or DVD) on which the archival information is stored. A library or consortium should determine the cost of the archive before signing a contract as a very high price may be set to discourage the exercise of this option.
The duration or length of the contract should be spelled out. It may be for one year with renewals, or for as many as five years. If the duration is one year with renewals, it must be clear that the terms set forth in the initial contract apply to the renewals-and for how many renewals.
At a minimum the eligible users should include all registered borrowers, not just those accessing the products from within a library facility. Ideally, anyone using a library-owned machine should have access, even those who are not registered borrowers. When negotiating this point, it is necessary to disclose the pattern of use by persons not registered because vendors--most of which use the number of registered borrowers as a factor in their pricing--are aware that reciprocal borrowing privileges and statewide borrower cards may significantly impact use of their products within a library. It is, of course, possible for a library to register all their guest borrowers.
The vendor must be assured that it is receiving income for each user of its products. Except as a transactional access (also known as pay-per-view or pay-per-search) agreement is in place, it relies on an authentication mechanism that qualifies each user before granting access. While most have their own authentication tools, they are willing to accept that of a library because they recognize that a library would like to avoid multiple authentications for multiple products. The contract should, therefore, identify the authentication product to be used and agreement by the parties that it is suitable.
Libraries and vendors use statistics for different reasons. Both will use them as a factor in determining price, but for libraries it is equally important to determine that they are getting value for money. They must be able to determine that the number of uses divided into the purchase price results in a reasonable cost per use. A library that negotiated for a comprehensive statistical package used the statistics it obtained to determine that one database was costing it $.19 per use; another cost it $6.30 per use. The average was under $1.00. It then launched a program to promote the databases that cost more than $1.00 per use. Some of the ones that remained expensive due to low use were subsequently cancelled.
Statistics on the use of individual titles in a package of journal titles are very important if the contract allows the library to remove subject categories or individual titles.
Among the features which a library may want to seek are customized links to the patron access catalog or to interlibrary loan request forms. While some aggregators charge separately for these, it often is possible to negotiate for them as a signing incentive.
Another negotiable feature is "branding," the tailoring of the look of the screens to match a library’s Web site and/or patron access catalog.
Most producers now have their own standard licenses. When a producer’s license applies to the licensee that is accessing the product through an aggregator, the producer license should be referenced in the contract between the library and aggregator. Typically, producer licenses stipulate how a product may be used, not its price.
Most vendors now use the number of registered borrowers as the basis for pricing annual subscriptions for online access to their products, but enrollment is a popular alternative for academic and school libraries and the number of employees for special libraries. Public libraries should strongly resist pricing based on the population served because patterns of use vary dramatically depending on a number of demographic factors. Total library budget, which is used by some vendors, should also be resisted because there is no correlation between it and the amount of remote database searching a library’s patrons undertake. Equally inappropriate is collection size.
The number of registered borrowers is the most acceptable of the options for public libraries. If that is the basis for pricing, it would be highly desirable to periodically renew patron registrations so that borrowers who have moved away do not inflate the subscription prices.
When negotiating price, it is essential to include credits for any print equivalents to which subscriptions are maintained. Many producers protect their print products by adjusting the prices of online subscriptions for those maintaining dual print/online subscriptions.
A subscription is not always appropriate. A library that expects a low level of use because a product is of potential interest to only a small number of patrons, may want to seek a transactional access or "pay per view" basis for pricing. If this is done, there should be a provision for changing to a subscription basis when, and if, the level of usage warrants it.
Yet another option is payment based on the maximum number of concurrent users that may access a product. This is a very equitable approach, but its suitability depends on whether the parties agree on the definition of "concurrent users" and there is a good mechanism for measuring the number of concurrent users. A public library consortium that negotiated a license based on the number of concurrent users discovered that the reason the vendor’s count differed so much from their own was that the vendor was counting log-ons and assuming that each would remain logged on for at least ten minutes. When challenged, the vendor revealed that it had no way of counting concurrent users.
Even when there has been agreement with regard to the impact on pricing from "content creep," there remains the typical annual change in pricing attributable to inflation. Few vendors will commit to holding prices for more than two years, therefore, it is important to include an "escalator clause" in the contract, a clause that caps the maximum percentage increase a vendor may add to the contract each year. The percentage typically is three to five percent.
It may become necessary for a library to cancel a subscription before the end of the contract period. Unanticipated cutbacks in funding are the most common reason. A library should negotiate a "non-appropriation clause," one which provides that cancellation on 90 days notice is without penalty if the library has sustained a budget reduction.
It is a lot simpler to contract with a limited number of aggregators for access to online databases, but a library that is willing to commit the additional staff time to the negotiation of multiple contracts may save thousands--even tens of thousands-of dollars a year.
A negotiation does not always result in the best possible terms. It is, therefore, a good idea to compare the results against the terms that are offered by a regional network or consortium. Many of the OCLC-affiliated networks and library consortia have negotiated agreements on behalf of their members with aggregators and individual database producers. Sometimes these agreements are the best available option.
Prepared by Richard W. Boss, Information Systems Consultants Inc.