Books (Photo credit: henry…)

On the face of it, patron-driven acquisitions (PDA) would appear to be a wholly negative innovation from a publisher’s point of view. Heretofore, libraries have purchased books just in case they were needed, but with PDA, the purchases are made just in time — on demand. This means that some books are never purchased and that those that are purchased may not be purchased immediately after publication, since a patron’s request may take place years after the date of publication.

Let’s work through the financial impact of PDA. I will restrict my remarks to the university press sector, though much of what I am asserting here is true for all book publishers.

Before we can estimate the impact of PDA on book publishers, we have to have some idea of what the baseline revenues are now. This brings us to the much disputed question of how many books university presses (or any publisher, for that matter) sell to academic libraries. I have discussed this question on the Kitchen before, and won’t repeat here all the reasons that determining sales to libraries is guesswork. My guess: library sales for university press titles come to 25% of the presses’ total volume. That’s 25% of $320 million, or $80 million. The figure of $320 million is an extrapolation based on AAUP figures, the 25% figure is a consensus number derived from countless conversations with university press staff and representatives of the book supply chain. The thing to bear in mind is that since the presses are not identical to one another, average figures can distort the measurement of presses that are outliers in some respects. For example, a press with a large trade program will have a lower percentage of library sales than a press with no trade program, which would be more dependent on library sales. The diversity among the presses means that PDA will affect them unequally.

If libraries were all to move to across-the-board PDA programs and no patrons requested titles, then the sales of the presses would decline by about $80 million, the amount currently being sold to libraries each year. But, (a) few if any libraries at this time are contemplating moving to an all-PDA program, and (b) press titles do in fact circulate in libraries, so we can assume that some portion of that $80 million is not at risk. What is at risk are the titles that never circulate. It’s commonly bandied about that 40% of books in academic libraries never circulate (an oversimplification, but useful for this analysis); if that’s true, then the presses are at risk for 40% of $80 million or $32 million. That figure represents a worst possible case, but the reality will not be nearly this severe.

In addition to the 40% of books that don’t circulate, however, we have to take into account the fact that of the 60% of the titles that do circulate, they don’t all get requested by patrons on the date of publication. Circulation varies by title, by library, by field, by so many things, but one analysis, the Walker Report, of the circulation records at the Cornell University Library, demonstrated that the period of circulation for a book is generally about 12 years. Thus a book published in 2000 would generally stop circulating by 2012. The circulation is not even over that period — that is, we don’t see one-twelfth of the circulation each year for 12 years — but still it would take 12 years for the publisher to receive all the income.

Here is an illustration. We have a hypothetical title that will eventually sell 1,000 copies to libraries (a very good sale for an academic book, but the virtue of that quantity is that it makes the arithmetic easy). We will assume (for the sake of simplicity) that the same number of copies are sold each year over twelve years. That comes to roughly 85 copies per year. In Year One, 85 patron requests result in 85 sales; in Year Two, 85 requests bring total sales to 170 copies; in Year Three, the total comes to 250; and so on up to approximately 1,000 copies in 12 years. If the books had been purchased when first published, the publisher may have received all the money in Year One, but with PDA, the receipt of the money is significantly delayed. So this makes the worst possible case referred to above (an exposure of $32 million) even worse. That twelve-year delay is a one-time occurrence — that is, after 12 years, sales through the just-in-time PDA system and the traditional purchasing method catch up with each other — but few businesses can survive such a shock to their cash flow.

The primary offset to these dire figures (from a publisher’s point of view; librarians may feel differently) is that libraries are not in fact going to all-PDA solutions. PDA is being tested and reviewed, tested and reviewed. While it is likely that PDA will grow significantly in the years ahead, from a publisher’s point of view it is being phased in. For many libraries, probably most, PDA will become just one of several ways libraries add books to their collections, so only a fraction of the 40% of books that don’t circulate are at risk. It will take several years of experience to know what that fraction is. In the meantime, publishers will be looking for new ways to sell books to academic libraries, a topic worthy of a post of its own.

Enhanced by Zemanta
Joseph Esposito

Joseph Esposito

Joe Esposito is a management consultant for the publishing and digital services industries. Joe focuses on organizational strategy and new business development. He is active in both the for-profit and not-for-profit areas.


17 Thoughts on "The Financial Impact of Patron-driven Acquisitions on University Presses"

As always, a thought-provoking analysis, Joe. Thanks.

To come up with practical strategic responses, I think it would be worth a press’s effort to refine your model to take into account the following:

1) Presses should have an average sales-life curve that they apply to their projects: e.g., 60% sold in year one, 30% in year two, and the rest over years three to five. That average, of course, has evolved in the context of informal PDA, which raises the third point for refinement. In the meantime, they could adjust this curve to see the impact of PDA’s phasing in.

2) Presses should also have an estimated sales ratio for ongoing frontlist/backlist sales. That “macro” forecasting tool will also have to adjust to accommodate the impact of PDA.

3) ARL and ACRL should be quizzed for stats on the behavior patterns of acquisitions librarians in the current informal PDA context and also for best guesses on the first-year impact of more rigorous PDA. Many scholarly publishers offer their monographs in series, and once a series has taken hold, libraries have been willing to set up standing orders, which raises the third related point.

4) Most presses should be able to identify on their own or with the help of library retailer/wholesalers what rough percentage of sales still occur through this admittedly dying sales channel.

5) A tougher point of which to take account is that libraries are considering applying PDA to journals, and that represents a significant downside to all scholarly publishers. For modeling that, presses might look to stats from ARL/ACRL (if they have them) on the rate of discontinuation of journals, although the reasons for discontinuation may vary from “lack of patron interest”. They might also look to engage more with document delivery services (or their own rights income records) to determine what the baseline per journal is for individual article requests, which could yield a very rough “worst-case” sales factor if libraries were to shift to a predominantly PDA-driven policy for journals.

As a whole, university presses would do well to engage with the library community to determine whether a subscription model for monographs can be worked out. Sandy Thatcher, Frances Pinter and yourself have suggested future models for monographs. With the shift from print to digital accelerating and PDA’s heaving over the horizon, it’s time to evaluate those models seriously and urgently.


Joe – this is a very good start to analyzing the impact of PDA on our business, but you’re leaving out one very important part of the equation – Short Term Loan.
The PDA model that seems to have been adopted by aggregators is to allow 4 short term loans at 10% of list price each before a purchase is triggered.
A book which may have only circulated once used to have the same value to the publisher as a book that circulated 10 times. This is no longer true when STL is part of the model.

I look forward to further posts on this topic.

I may be wrong, but an enterprise producing items no patron orders is not really a business.

Tall Timber, you have hit the nail on the head. What right has any business to stay in business if it relies on (largely) state funds to buy stuff that nobody uses? I read the Against the Grain discussion last June and whilst I like the idea that the state (through the libraries) support books that are not financially viable? If PDA is a more honest system then maybe publishers are right to be fearful.
BUT now that I have been fortunate enough and had the opportunity to discuss PDA with the libraries in Australia who have been doing this for a long long time is that their book expenditure has gone up. Significantly up. So much so that journals expenditure has been reduced as the libraries buy more books. There is an anecdotal rule that there are far more books needed and wanted than a library can afford to buy each year and PDA appears to make those books available. To go directly against Joe’s opening gambit that PDA is a, “…wholly negative innovation from a publisher’s point of view” contradicts the evidence.
I think that whilst I appreciate Joe’s attempt at sizing and assessing the finance of PDA the math appears to be significantly flawed to me and it appears to be almost scaremongering. PDA is nearing maturity as a model in certain parts of the academic world and it seems to produce a healthier balance for library, patron and publisher.
To read these articles you may think that Chicken Licken was at the head of UP ebook research. The sky is not falling in on our heads. If we look beyond these rather insular 50 states we can see that PDA actually works pretty well for most in the chain. If they have a viable business of course. Which, as Tall Timber states, you may not have if nobody actually reads your books.

There are clearly words missing from my opening paragraph: “I read the Against the Grain discussion last June and whilst I like the idea that the state (through the libraries) support books that are not financially viable?”
I like = the tax payer supporting pedagogy and worthy academic narrative whether financially viable or not
I don’t like = the fact that it must be supported so inefficiently through insisting libraries pay millions of dollars to house cloth books that nobody opens

If 40% of your product never leaves the shelf, do you really have a business?

Thanks for the post, always interesting to see the publisher’s point of view. Are you thinking about libraries buying ebooks, or print, or both, using PDA? With ebooks there will be no need for libraries to purchase multiple copies or to replace damaged or lost copies, which will further deplete sales I’m afraid. I’d suggest that in the past ten years or so librarians have become much more efficient in their purchasing, mostly because journals have taken up so much of their budget. So it’s a matter of the publisher/supplier/library to negotiate a fair price through PDA which represents the costs of producing the text. [Or for academics to publish through Open Access?].

I think Joe’s analysis is right on target for the most part, though I am glad that he recognizes that the impact of PDA on presses with different types of lists will vary greatly. What about presses that publish a lot of regional titles, for instance? How will PDA affect them, if at all, since probably only libraries in the state will order titles having to do with that state, if they order them at all, and those that do are more likely to be public than academic libraries (though some regional titles are also solid works of scholarship). One category of book that may do better under PDA than approval-plan ordering is the revised dissertation, since patrons do not care about a book’s origin in the way that libraries with subscriptions to ProQuest do. Also, it must be emphasized that, without digital printing (SRDP and POD) the effect of PDA would be a whole lot worse because a press would have to carry inventory for a much longer period of time if the buying cycle stretches out to a dozen years or more.

Nevermind the publishers…. If 40% of what one buys for the shelf is never used, does the library have a good operational model?

But I’d argue that in many libraries the percentage is very much lower. (See previous comment). I’m writing from UK, not sure if this makes a difference.

In some libraries it’s certainly true that the percentage of unused books is lower. But as Tom suggests, the model itself is a problem — it’s based on third-party guesswork about future needs, and is inevitably both inefficient (in that we often buy the wrong books) and ineffective (in that we often fail to buy the right ones, as evidenced by ILL statistics). If your library has very low rates of non-use and if there is very little demand for your ILL services, then PDA is probably less of an urgent issue for you.

Agreed. Honestly, I think that it all comes down to mission. If you are at a flag-ship institution with a long-standing tradition and understanding that you collect as a resource center for others in your state or region, you would likely find PDA meeting a smaller percentage of your overall mission. That’s not to say that it is not a part of your mission. It is, and it should be. But, it’s likely to be a smaller percentage of the total when compared to your firm orders, approval plans and blanket orders, etc….

I like to equate PDA to a transportation option. Few technologies completely replace their predecessors. In the case of transportation, your community may find occasion to benefit from foot transportation, cars, busses, bikes, airplanes, horses, etc…. They all fit into part of a bigger picture. The challenge is finding the right balance that is appropriate for your community.

That said, I would still like to see more than 60% of the material we buy used, especially for the more commonly available materials from North America and Western Europe.

I don’t think you can simply look at current sales and then subtract titles that won’t be bought under a PDA model. Under the traditional model, a new book is purchased, so users have access to one more book. They might find it useful; they might not. For small libraries in tough budget times, very few books are added in a year. However, if you start up a PDA program, you might suddenly gain access to many thousands of new books, depending on how the plan is structured. Now, it’s very likely that the students and faculty will find a useful book.

I think that PDA will prevent libraries from buying books that don’t circulate. However, it will also cause libraries to buy books they never would have selected. In our recent pilot, our users accessed many books that subject librarians would not have chosen. I think it is quite conceivable that libraries would spend just as much. We’ll just be spending it on books that get read.

Comments are closed.