The Logical Extension of the Demand-Driven Purchase Model: Customization, Multimedia and Ever-Improving Analytics
by David Parker
Demand-driven acquisition (DDA) in the ebook space has been with us long enough to have generated both acclaim and reproach. New publishing initiatives to create demand-driven friendly content have emerged, untethered from legacy print-based pricing logic, in parallel with the inevitable backlash that has seen publishers pull content from DDA distribution because of usage too low to trigger purchase. Journals have been slower to the DDA game, although options are emerging through platforms like Read Cube and article rental programs. And video has been available in DDA and Evidence-Based Acquisition (EBA) for several years and demand is growing at pace. Librarians praise the return-on-investment (ROI) when lightly used content gets read, viewed, used but not in sufficient volume to trigger purchase and only very high-use content is triggered for purchase. Publishers and aggregators of content enjoy the “long tail” opportunity to expose their back catalogs at very low marginal cost, but worry about the financial impact on their front list should it not drive enough views to trigger purchase.
We are in the very early days of DDA across media types and a fair amount of sorting will take place in the coming years as we establish equilibrium between library purchasing efficacy (ROI) and proper compensation for content creators to sustain their enterprises. I see three key points upon which the future of DDA will evolve and, in doing so, will bring us closer to a market-responsive equilibrium between the needs of content providers and the demands of content consumers:
1. Customization of trigger, price, and length of trigger view and period/scope of access to content post-purchase.
2. Multimedia DDA platforms that include text, image, video, and audio all in one platform.
3. Data analytics that empower both the library and the content provider.
There really is no natural limit on a specific library’s possible DDA profile. Given a sufficient progression in the technology of the publisher or aggregator-provided platform, each library account should be able to customize its experience.
Four inputs come together to form the “fulcrum” of negotiation between the content provider and the library: the price paid, the length of the view that triggers the purchase, the number of views that trigger the purchase, and the period and volume of access once a DDA purchase is triggered. The “fixed” view we have on this today is hurting all of us as content and value-in-use of content in the library is far too diverse to be captured in a single model.
Further, different libraries and different librarians will value the same content differently. I often come back to the example of a classic ethnographic film. Such a film is a staple of an Anthropology 101 course and can be viewed by hundreds of students a semester, but only “viewed,” from the perspective of a DDA trigger, once a semester. This film is highly prized by the creator and by the consumer, but the current DDA model fails to capture this value as it only measures aggregate click-throughs or hits.
Imagine, instead, that the library could gain access to this film, and many more like it, with the following DDA profile: Purchase for perpetual access, multi-viewer, on the first view that exceeds five minutes for a price of $750.00. Then couple this profile with a separate profile for all other video that states: Subscribe for one-year access, multi-viewer, on the third view that exceeds 30 seconds for a price of $99.00. You can extend this logic to high-use e-textbooks versus esoteric scholarly manuscripts, or high-use audio tracks for music appreciation courses versus obscure performances on the Clavichord. And on and on.
Without such a model, much high-use, high-demand content is likely to not be available in DDA.
Why should a librarian or library patron have to access DDA content via multiple platforms and vendors?
Ebooks, archival documents, audio tracks, video, musical scores, data sets, journal articles, etc. are all content types amenable to measurement, use, and sale. The aggregation and distribution of content by small and large companies alike is increasingly multimedia and, therefore, the platforms must eventually also be multimedia. Specialized collections, especially in areas like music and film studies, provide scholars and students with a mix of media types for study.
Of course, the “Modern French Film Studies Collection” can be purchased, and the individual items within the collection can be purchased via single title sales, but DDA via a single, multimedia platform allows usage to determine the purchase pathways and the student of modern French Film might well be the trigger of the purchase of a video, a film script, a reference monograph on the film and a biography of the director…or not, depending on the level of interest and the purchase trigger parameters selected by the librarian.
Massive aggregation of ebooks, through platforms like Ebrary, delivered DDA, has allowed libraries to migrate toward a single ebook platform and evade, even partially, a state of “platform weariness.” Imagine then a future state where a single mixed media DDA platform supports access to all the media types central to a student or scholar’s search within and across disciplines and areas of study.
Seven years ago, when I founded Business Expert Press, our ebook collection was made available exclusively through the ebrary platform. At that time, Counter statistics provided little more than title-level and collection-level numbers of views and total pages viewed; neither we at BEP nor the libraries that purchased our collection had visibility into usage beyond these raw statistics. I suspect our internal team spent more time reviewing usage statistics (to assess the likelihood of a given library continuing to purchase our collection) than did our library customers.
Usage data, to be truly valuable, must be more robust. Near the end of my time at Pearson Education, the company acquired eCollege. I recall the standout feature of eCollege that made it such a desirable acquisition target for Pearson was the robust “back office” data they provided university administrators on online course, program and instructor efficacy; measured both in student results and program profitability (i.e. enrollment rates and completion rates along with costs to support a given class). The state of data analytics in support of the classroom and the learning enterprise has become increasingly individualized to the learner and has moved forward at a much faster pace than within the library content space.
Fast-forward to 2015 and the current state-of-play in DDA. At present, we are giving librarians little more than data about pages read, minutes viewed, total views, total users, etc. What if a librarian were able to distinguish between the views of faculty versus students? Ph.D. students versus undergraduates? The type of content (learning as opposed to scholarly reference) viewed by which departments at which time in the semester? Discipline profiles? Correlations between online programs, degrees and content triggered for purchase? Location of view: classroom, dorm, in library, off campus? Perhaps some of these possibilities strike a chord with you the reader and, perhaps, some of this is improbable and unnecessary. But evolving usage and user data that offers ever deeper insight into the values and needs of the library patron is the indispensable corollary to the description above about customization of the DDA profile: a library that knows a good deal about how its content is used will make ever better decisions about how to trigger purchase in an improved future-state DDA environment.
It is my hope that this column sparks debate, inspires publishers and aggregators and sets off a conversation about how far and how fast we move with DDA.
David Parker is the Senior Vice President of Editorial, Licensing & Marketing at Alexander Street Press.