The simplest book-buying model in the library world is one where a library buys a book from a publisher, the contact between the seller and the sold product becomes severed, and that’s the end of the transaction. This is the same model that operates when, say, I buy eggs from the market. With the advent of e-books, however, cost models have become much more varied and interesting.
The fundamental technological change explaining the explosion of e-book business models is that publishers can now “sell” an e-book and still remain attached to their sold product. That is, most e-books “sold” remained hosted on the publishers’ — or their preferred agents’ — servers and hence under the sellers’ control. This allows a myriad of factors to come into play. Publishers and libraries can now negotiate over the conditions of access that libraries get with their purchases. How long will libraries get access? Who from the libraries can access the content? In how many ways can different access conditions be combined into unique packages? One imagines an almost infinite number. Below I’d like to highlight a couple of innovative cost models that I came across recently.
One of the basic innovation that emerged with e-book selling is patron-driven acquisition (a.k.a. patron-initiated acquisition or demand-driven acquisition). Here’s how one e-book vendor, EBL, describes its demand-driven acquisition model
Using EBL’s Demand-driven Acquisition tools, you can catalog and display thousands of ebooks that you haven’t bought. Patrons are provided with a free preview of the full-text and then either allowed further access to the title either on an automated or request basis. Libraries can choose to automatically purchase those ebook titles after a specified number of uses, or pattern of demand has been established.
From there you can already see where the multiple branch off points happen that can result in multiple different cost models. For instance, what are the patterns of demand? Possibilities abound. An acquisition can be triggered after any specified number of uses, from 1 to however many (with some upper limit, I suppose). What counts as a use? Any clicks on the content? A click where the user lingers for at least 5 minutes? 10 minutes?
A newest thing constituting a “pattern of demand” that I know goes by the name of demand-proven acquisition. In this model, a library commits to spend a certain amount of money with a publisher, say, $20,000/year. In return for putting down this non-refundable deposit, the library gets access to, say, $100,000 worth of content. All of this content is then made available to the library’s users (but remains hosted by the publisher so that content can be removed at anytime). During the course of the year, library patrons can use any of that $100,000-worth of content, and at the end of the year, the library chooses which books from this selection to keep permanently. The library may choose to buy the entire $100,000 collection, or it may choose to buy only a portion thereof, but it must buy at least $20,000 worth. (What the library chooses not to keep, library patrons lose access to after that one year.) Because libraries will presumably buy only content that has already exceeded its own set criteria of sufficient use, this model is known as “demand proven” acquisition. The model essentially differs from the basic patron-driven acquisition framework in that libraries get to choose ex-post which books to keep instead of agreeing with the publishers beforehand on when acquisition is triggered.
A variant from this demand proven acquisition model is for library to put down a deposit of, again say, $20,000 to get access to $100,000 worth of content. After the first year of use, library then gets to return a specified fraction of the titles. If it chooses to return nothing, then the library pays another $20,000 for another year worth of access to all the titles. If this continues for 5 years, then the library will have purchased the entire collection and can now exert complete control over the content, severing the link between the product and the publisher if it wishes by hosting the content on its own site. However, if the library chooses to return titles during any year in that first five year, then it will pay less than $20,000 to the publisher that year. At the limit, the library may choose to return the entire collection after the first year, thus effectively having rented a $100,000-worth collection for a year for the cost of $20,000. Most contracts, however, exclude the possibility of the library returning the entire collection by putting a cap on the number of titles that may be returned. The cap can then be made to vary with number of years the contract has run (e.g. 10% of the titles in the first year, 15% in the second . . . etc.).
Even in this broad sketch of the e-book cost models, you can already see how many details may be changed so arrive at a different model. Isn’t that fun? Of course, there must be a balance struck between variety/choice/flexibility and complexity. For somebody who’s stumbling onto these details for the first time, however, the variety sure is neat.