Publisher feedback on PDA models

1. Introduction

1.1 Executive summary

This report compiles the responses from e-book suppliers to questions on patron-driven models of e-book acquisition (PDA) for consortia. These questions were put to them in a series of telephone interviews conducted by JISC Collections staff in January 2013 as part of the E-BASS 25 project. All responses are unattributed at the request of some of the participants. The report concludes with a series of recommendations for institutions considering using PDA as a consortium, derived from the responses received.

1.2 The interviews

The purpose of the interviews was to elicit the opinions of the suppliers (who included publishers and aggregators) about the four PDA models being considered by the project, and how these would work as a consortium purchase. The aim was to discover which they thought were feasible or realistic in this context, and to learn their thoughts on the practicalities of how these models would operate. The intention is to use this information to construct the guidelines for institutions considering embarking on a PDA consortium project.

1.3 The process

JISC Collections staff (Ben Taplin and Carolyn Alderson) interviewed seven publishers and two aggregators. The publishers included small, subject-specialist companies and large, multi-disciplinary organisations. Apart from one, which was answered by email, the interviews were conducted over the telephone and lasted approximately forty-five minutes each. The list of questions, which had been drawn up in consultation with the E-BASS 25 Project Management team, was the same in each interview (see Appendix 1), and was sent to interviewees beforehand, along with the outline of the four models. Responses were collated and are organised thematically below.

1.4 The models

Model 1 is a purchase model, whereby the library is supplied with a collection of e-books, which are made discoverable to the library user. After an agreed amount of “significant use” of a title by a user, or “patron”, a purchase trigger is activated, and the book is acquired permanently by the library. Model 2 is a rental model: similar to Model 1, but the significant usage triggers a paid-for rental period, rather than an outright purchase.

In Model 3 the library is charged for each significant use of an e-book, up to an agreed amount, after which the title is owned outright. At the end of the agreement, access to titles that have not reached their usage limit is lost. Model 4 offers access to a collection of e-books for an agreed period, after which time the library decides which titles to acquire permanently, based on the levels of usage each has received. This is known as an “evidence-based selection” (EBS) model, and is not a purely PDA model, as it allows a degree of intervention on the part of the library. Usually the amount of money the library will spend at the end of the period is decided and paid upfront at the beginning of the agreement.

A fuller description of each model is available.

1.5 The consortium

The hypothetical consortium the suppliers were asked to consider comprised academic institutions of various sizes with different subject interests, differing levels of research intensity, and independent local budgets.

2. The responses

2.1 Thoughts on the models

The interviewees were asked which of the four models was most appealing to them, which they thought was fairest for a consortium, and which they did not like.

Perhaps unsurprisingly, the evidence-based selection (Model 4) proved popular with the publishers, as it would be relatively straightforward to implement, and guarantees them some upfront payment, helping their sales forecasting and planning. Seven responses favoured this model. One publisher likened it to a “paid-for trial”, and one suggested that the upfront payment would be an incentive to the institution to ensure the content was used. Others mentioned its greater flexibility and sustainability. Although, according to one publisher, direct PDA models may seem initially attractive because the absence of library control could bring in higher revenue, the EBS model involves a higher level of trust and cooperation between both parties and is therefore “more interesting in the long term.”

Models 1 to 3 were generally seen as being complicated to manage for the publisher or aggregator, entailing a considerable amount of technical work with no guarantee of payment if content is not used. This complexity was seen to increase if micropayments from a large number of institutions were involved.

Those who expressed a preference for the more directly patron-driven models (1-3) pointed to their relative fairness, in that libraries are paying only for content that is actually being used. One of the largest publishers stated that Models 1 and 2 are reasonable, as they reflect usage, and so justify the purchase: no publisher wants to defend purchases that are not used. One of the aggregators, who already offers Models 1 and 2, preferred them as they are the easiest to scale from individual institution accounts to accounts for consortia.

Only one respondent favoured Model 3, mentioning that its direct link between level of usage and price made it the most accurate and quantifiable model.

Figure 1: Preferred models of PDA

2.2 Which models are currently used?

Although the EBS model was so popular, it was interesting to learn that only two of the organisations we spoke to currently offer it. Three use a system closest to Model 3, and two offer Models 1 and 2. Most of the publishers mentioned that they would be happy to offer Models 1, 2 or 3 through a third party aggregator, even if they themselves don’t currently have the technology in place to offer them directly. Most publishers are willing to be flexible to some degree if new sources of income are being proposed.

Figure 2: Currently used models of PDA

 2.3 What constitutes “significant use”?

Definitions of what level of usage is deemed significant enough to trigger a payment vary considerably across the respondents. Whereas one aggregator offers only tables of contents and abstracts for free and charges for any further use (“opening the book” in their words), the other allows ten minutes of consecutive use of a single title by a single user, or ten pages read. One of the larger publishers has similar parameters, and has allowed two of these significant uses before a payment is triggered in some agreements.

Those publishers already using the EBS model provided standard COUNTER-compliant data to libraries, and a “significant use” was deemed to be any use of the content. The librarian can then make a decision about what to purchase, and may use other criteria besides usage.

There is clearly room for some negotiation around payment triggers, although this particular publisher was keen that the definitions be kept hidden from end-users, as they claim to have found in the past that students will manipulate the system to avoid payment.

2.4 Accommodating specialist institutions

In terms of which model was seen as fairest for a consortium that included institutions with specific subject interests or local considerations, two suppliers again mentioned Model 4, as it allows for some mediation or control of spending on the part of the institution. Others were of the view that special interests would be difficult to accommodate in a consortial agreement. One of the aggregators recommended any specialist institution having a personal account as well as belonging to the consortium.

2.5 Concerns over PDA

General reservations were expressed about the potential for each model to offer access to content for free. In Models 1-3 this means usage for short periods of time, below the payment threshold. This is pertinent for specialist publishers whose content receives briefer use. In Model 4, the concern is over the large amount of material being made available for no payment, especially for those publishers without DRM on their content.

The threat posed by e-books to print sales is still a concern to publishers, especially those in the textbook market, despite lack of firm evidence for any correlation. This threat is seen to be magnified by any consortial agreement in which single copies of books are shared across several institutions. This could mean that some publishers limit the titles they make available.

The idea of single shared copies also causes difficulties for payments of royalties to contributing authors: not a problem that might occur to librarians perhaps, but still a very real concern for publishers. One scientific publisher did however emphasise that the possibility of their authors’ research being more widely disseminated through such an agreement should also be viewed positively. One of the larger publishers circumvents such problems by not offering shared copies: instead each institution receives its own copy, thus also addressing questions surrounding ownership if the consortium breaks up.

3. Managing consortial PDA: the practicalities

Most of the organisations we spoke to about the day-to-day practicalities of implementing and managing PDA agreements for consortia were confident that they could overcome any problems, although many expressed a wish to keep arrangements as straightforward as possible. For some of the publishers, any agreement of this nature would have to be managed by a third party aggregator.

3.1 Licensing

All respondents were happy to use a model licence to reflect the libraries’ requirements for such an agreement, instead of their own licence. One of the larger publishers, however, would prefer not to use the JISC Collections model licence, which was seen as a document that favoured the institution over the publisher, and was used to push for further rights.

3.2 Payment

Although most of the publishers showed flexibility in their arrangements for ordering and payments, giving consortium members the choice of ordering and paying individually or as a group, both the aggregators preferred to centralise these functions. The larger of the two aggregators will only handle consortium orders (and in fact all communication) through a single dedicated contact, which could be problematic in terms of staffing, and so on.

In terms of the multiplying factors or discounts that publishers offer to consortia, these tend not to be fixed, and offer scope for negotiation on the part of the consortium.

3.4 New editions

The consensus among publishers is to treat new editions of titles as new items that require a separate purchase.  Some publishers delay the release of new titles in electronic format so as to minimise the harm to print sales. When a title that is part of an agreement changes publisher, most of our respondents confirmed that they would be able to honour their commitments. If a publisher loses the rights to a title that an institution has paid for, a replacement title of similar value or a refund would be offered instead.

3.5 Standards

None of our respondents foresaw any trouble in supplying usage statistics to the consortium members at a group or individual level, nor any problems caused by members using different authentication methods for access. All provide free MARC records, and nearly all of them supply information to the commercially available knowledge bases and discovery services, such as Summon, and Primo. All comply with accessibility standards to varying degrees.

4. Conclusions and recommendations

The suppliers we spoke to are currently at very different stages of being able to offer all the models presented by the E-BASS25 project, but were united in their willingness to explore the practicalities of each model, and to be flexible in order to accommodate the requirements of a consortium. All of them expressed certain reservations about the various models, and about consortial agreements, however if the consortium were able to allay these concerns, then they all offered potential negotiation room for institutions, in terms of pricing, and definitions of significant use.

From the responses we received, it is fair to conclude that suppliers will be willing to offer more favourable terms to a consortium that can comply with the following requirements

4.1 Ordering and invoicing

The process for ordering and invoicing should be as straightforward as possible. Models that entail micropayments should be avoided, and if possible, the option of a centralised ordering system should be offered. In order to facilitate budgetary planning, careful thought should be given to the dates when the institutions are to be invoiced, and these should be confirmed at the start of the agreement.

4.2 Participation

There should be a firm commitment about numbers and sizes (FTEs) of participating institutions. Institutions leaving or joining the consortium during the term of the agreement should be avoided where possible, but a policy for how this eventuality should be handled should be in place at the start of the agreement. If discounts are offered on a sliding scale depending on levels of participation, then a cut-off date needs to be agreed.

4.3 Content

There should also be firm agreement about the content that is being made available. Publishers are wary of making large amounts of their content available without payment, and so an agreement that works within the parameters of a proscribed, pre-agreed list would be viewed more favourably. Make sure the supplier has a robust policy for honouring the addition or removal of titles from the list.

4.4 Expenditure

Overall, greater levels of commitment on the part of the consortium are rewarded by greater discounts from the provider. The commitment most valued by suppliers is an agreed level of expenditure. If this can be built into the agreements (Model 4 is the most suitable for this) then the provider is likely to be more generous.

4.5 Licensing

All obligations, of all parties, should be agreed at the outset, and signed by all participants. If using an EBS model, in which access to some of the content will be lost at the end of the agreement, careful consideration should be given to the licence’s definition of “Licensed Material” and the permitted uses of the content: in particular the archival rights to any copies made of content that is not retained at the end of the agreement.

Appendix 1: questions to suppliers

  1. Which pricing model is most appealing/offers most opportunities to the publisher and why?
  2. Which of the models are used by the publisher already?
  3. If you currently have a PDA model – what is the “significant use” trigger for purchase? – i.e. 3 views, 5 downloads etc?
  4. What pricing model would be considered most fair for a consortium with different subject interests or local considerations?
  5. What don’t you like about any of the pricing models as a consortium purchasing model?
  6. What are the risks/drawbacks to the publisher in using any of the models – explain?
  7. Would you be willing to use a model licence (such as the JISC Collection’s Model Licence) as the basis of the agreement, rather than the publisher’s own licence?
  8. What would the publisher want to be clear about in terms of commitment from the participants?
  9. How would you expect to manage the order and payments process?
  10. How would you deal with a library joining / leaving the consortium after the agreement had started?
  11. How do you typically deal with new published editions for a consortium agreement?
  12. How do you typically deal with titles transferred (in and out) in a consortium agreement?
  13. Do you normally provide publisher MARC records?  Do you provide e-book metadata to commercially available knowledge bases? If not, how do you ensure resource discovery?
  14. How would you intend to manage a programme for a consortium where usage is aggregated across the participants etc, but where individual institutions have differing requirements in terms of authentication, access and discovery?
  15. What levels of accessibility do you offer for users who struggle to read print?
  16. Any other thoughts?

Questions about specific models

PDA Purchase model – what are the relevant multipliers that might apply when purchase is triggered, presumably on a sliding scale (e.g. X institutions at Y x Standard Price – see Arizona for example)

PDA Rental, PDA Usage

What usage stats should be provided, how often and how accessed in order to make the model attractive to librarians and administrators?

Evidence Based Model

If the consortium guaranteed a number of participants what kind of discount might be applied to encourage greatest participation?