Struggling to Satisfy Demand
the reality of the public library ebook marketplace reflected in usage data from a selection of public libraries
Matt Weaver, Board member, Library Renewal
In order to serve our constituents with electronic content, libraries need to be able to understand how our collections are being used. This paper aims to present library-centered usage data to help libraries make decisions with regards to e-content, and to counter media and industry hype. Much has been written about the impact of major publisher changes on library lending, which are noted in the Sidebar. By looking at these events in the context of actual usage data, this report endeavors to demonstrate that a vendor-driven ebook model is neither extensible nor sustainable.
Purpose and Scope
In the period from January 2011 through June 2012, there were many major events that changed how libraries have been able to provide ebooks and digital audiobooks. As a result of these decisions, the ebook/audiobook market is split between vendors and devices. Costs of content and administrative fees are growing. As aggregate “big data” on ebook usage abounds, not all of that data comes from the perspective of libraries.
This period has been critical in laying the groundwork for ebook wars among library vendors. With the market split between vendors who have exclusive access to publisher content, and/or support for dominant ebook format, libraries will feel pressure to add multiple vendors. Libraries that rely on vendor-generated data without looking at their local usage see a distorted picture of how ebooks are affecting libraries. This paper presents data from a small sample of public libraries to get a sense of how public library ebook and audiobook collections are being used. While this paper hardly qualifies as “big data,” the snapshot that it captures presents insights into usage that have not been shown in other reports, surveys or studies.
This paper presents data from a handful of librari
es that have provided statistical reports generated via Overdrive’s Content Reserve interface, and other sources. All reports were generated in the second-half of 2012, before Overdrive released the new version of this administrative suite.
For this report, only data for Overdrive was used. Why just Overdrive? While competition has increased in the ebook market, Overdrive was by far the largest player in the market during the scope of this analysis.
This report uses data from three members of Library Renewal’s Partner Libraries program:
- Robbins Public Library (MA)
- Santa Monica Public Library (CA)
- Omaha Public Library (NE)
Two other libraries, which are not partner libraries, also provided data:
- Topeka Shawnee County Public Library (KS)
- Westlake Porter Public Library (OH)
Notes about data sources
Topeka Shawnee County Public Library had been a customer of Overdrive via the State Library of Kansas consortium agreement. In December 2011, after the State Library had ended its agreement with Overdrive, TSCPL became an independent customer; however, the data that covers the period in which TSCPL was part of the consortium agreement disappeared. As a result, TSCPL’s data covers December 2011 through June 2012.
Not all libraries provided data for each area of analysis. Omissions are noted.
Disclosure: Library Renewal Board Member Matt Weaver works at Westlake Porter Public Library, and Board Members Gina Millsap and David Lee King work at Topeka Shawnee County Public Library.
Provision of data for this project does not imply a relationship with, nor an endorsement of Library Renewal.
Overdrive Big Data: An Overview
During 2012, Overdrive has released several reports. The first Big Data report, released first in April but was more broadly available in May, focused on patron engagement with Overdrive site content and contained no checkout data. Featuring d
ata on page views, cover image views, and site visits, this report’s audience clearly was not libraries, but rather publishers and the message was clear: library patrons engage with publishers’ content.
During the summer of 2012, the American Libraries Association sponsored a survey that Overdrive ran on libraries’ Overdrive-run websites. The report created from that survey includes some useful data regarding devices and sources of book recommendations; but the most touted takeaway from the report was the fact that library ebook borrows are also library ebook buyers.
In January 2013, the company released a report that stated that 70 million ebooks were checked out in 2012. Out of context, those numbers appear impressive; but when considered in the context of total library circulation, total ebook circulation is a fraction of overall circulation. In January 2013, the IMLS released data for U.S. public libraries covering fiscal year 2010: total circulation was 2.46 billion items in 2010 (Swan et al, 2013), up from 2.41 billion items in 2009 (Miller, et. al, 2011). If total 2012 U.S. public libraries’ circulation were even 2.4 billion, ebook circulation would amount to less than 3% of total circulation based on Overdrive’s numbers.
General usage data
It comes as a shock to no one who works in public libraries that the holidays represent a huge increase in new Overdrive users. In
2011, ereader shipments totaled 27.7 million units (IDC, Dec 2012) and tablet shipments reached 68.7 million units (IDC, Mar. 2012). With the increase in demand for devices came
In Figures 1 and 2, there are clear spikes in usage in December and January. The boost in new users in Figure 2 results in a larger number of unique users with checkouts following the holidays and adds enough users to the pool of active users of the service to cause an increase in monthly checkouts (Figure 4) in the spring months that remains higher than the fall months, despite the drop in the addition of new users in February and March (Figure 2).
When looking at each month’s checkouts as a percentage of total checkouts for the January 2011 to June 2012 period (Figure 3), the impact of the post-holiday boom is apparent. Some months in the spring posted higher figures, but that growth built on the rise in during the holidays.
One thing is clear: the acquisition of new devices drives usage. Nowhere is this more patent than in Figure 2, which plainly shows a surge in December and January. The events noted in the sidebar have radically altered access to content, and threatens libraries’ ability to provide enough content in order to keep them.
Data in Figures 1 through 4 reveals bumps after the arrival in the Kindle, in September; but the full impact of the Kindle on library ebook lending is revealed in data that shows the Kindle format’s share of checkouts, as shown below.
Formats, Publishers and Competition
|Table 1: Kindle share of ebook checkouts – average of sample libraries (TSCPL is not included)|
In September 2011, when the Kindle format was first supported by Overdrive, it accounted for an average 22.3% of ebook checkouts among the libraries that provided data for this category (Omaha, Robbins, Santa Monica, and WPPL). By December 2011, that share doubled to nearly 55%. In March, Kindle accounted for nearly 60% of ebook checkouts, falling slightly by end-June 2012, the end of the term covered in this analysis (Table 1).
Big Six Content
Table 2 shows the changes in major publishers’ share of total checkouts between January 2011 and June 2012. Despite having ended its agreement for library lending through Overdrive in February 2012, Penguin’s share only dropped by a little more than two percentage points. Random House’s share increased, despite its price increases which took effect in March 2012. The timing the Penguin and Random House decisions may have been too close to the end of the scope of this project to see them reflected in the data. Random House’s share increased by almost twelve percentage points. The share of Hachette, which only allows libraries to loan backlist titles, remained constant, but low. Harper Collins’ share dropped by more than half.
|Table 2: Share of checkouts – average of sample libraries (TSCPL is not included)|
|publisher||Jan. 2011||Jun. 2012|
|Random House, Inc.||
|Penguin Group (USA), Inc.||
|HarperCollins Publishers Inc.||
|Hachette Digital, Inc.||
The landscape for ebooks in libraries changes constantly. 3M and Penguin embarked on a pilot project for libraries shortly before making the publisher’s content available to all 3M Cloud Library customers. Hachette is working on a library lending pilot program–not with Overdrive, but rather its competitors, 3M, and Baker & Taylor. Macmillan has announced that it will work on a pilot ebook lending program with several library distributors, offering backlist titles from one of its units, Minotaur–however, not to consortia.
Given that reportedly, Penguin ended its relationship with Overdrive because of terms by which libraries received Kindle support, it is highly unlikely that Penguin and Overdrive will rekindle (sorry) their partnership, as long as Amazon is in the mix.
Other competitors promise some support of Kindle support, via apps that can run on the Kindle Fire series of devices; however, Overdrive is the only library ebook vendor that supports the much more affordable Kindle e-ink ereaders.
What are libraries to do? A library that has been a long-standing customer with Overdrive would lose all of their content should they choose to move to another vendor, no matter what Random House says about libraries’ “ownership” of content.
Table 3: Library Ebook Vendors
|Big Six Content||Penguin, RH, HC||RH, Hachette, HC||RH, Hachette (backlist), HC||no big six|
|Kindle Support||app only||app only||all wifi-compatible Kindle devices||app only|
If a library that is currently an Overdrive customer wants to restore access to Penguin titles, it will have to not only pay another administrative fee and content costs for the second collection, but in adding the 3M Cloud Library, it will introduce more confusion for Kindle owners.
I take you to a public service desk at Your Public Library, Anytown, USA:
Member: I would like to download a Nora Roberts book to my Kindle.
Staff member: I need to know what book and what type of Kindle you have, because if the book is new and you have a Kindle Fire then I will show you how to download the 3M Cloud Library app. If the book is an older title that we got through Overdrive, we can get it–but you will have to transfer it via USB cable. Even if it’s a Kindle Fire, and even though there is an Overdrive app for that device. If you have a Kindle e-ink ereader, and the book is new, we can’t get it for you at all. Have I lost you?
Yes. I’m guessing we will have lost them. With Kindle support split between device format, owners of e-ink Kindles are locked out of content from competitors.
If all of this doesn’t have current or potential Overdrive customers reaching for the antacids, there is still the concern that the Penguin-Random House merger could result in Overdrive losing Random House titles, depending on which half of the merged whole decides how the new entity will deal with library vendors. Should Overdrive lose Random House titles, then its service amounts to little more than a Freading with exorbitant administration costs.
At what rate will ebook vendor administrative costs increase? In 2011, the State Library of Kansas was faced with a 700% increase in administrative costs alone over the course of a 3-year renewal from $10,000 to more than $75,000. The mileage of different libraries–consortia vs. independent customers–is likely to vary. As for content costs, with increases of up to 300% announced by Random House, and another increase announced by Hachette, the systems and content agreements that control ebook content are preventing competition from having a positive benefit for libraries. The result is, less content for our constituents, and spending more of their taxes to keep what libraries already have.
Don’t believe the hype, nor specious logic
While there is no question that ebooks have become established as a mainstream format, many media pieces about ebooks in libraries are inflating their usage beyond the reality. Take, for instance, a piece that says that ebooks are outpacing print books within a county library system:
“In 2012, 19,948 ebooks were taken out from the [county] library system, compared to the circulation of 17,678 items at the Woodbine library branch and 16,598 at Stone Harbor, according to figures provided by Poillon.”
Comparing all ebooks for a county library system to the circulation data of two separate branches, isn’t sound reasoning with which one can conclude that ebooks are killing off print circulation at libraries. In fact, the data shows the opposite of the author’s conclusion: total circulation of those two branches alone–which must be small branches based on that annual circulation data–exceeds ebook circulation for the entire Cape May County. According to IMLS data, in fiscal-year 2010, Cape May County Public Library’s total circulation was 522,310. Without a 2012 circulation to compare accurately, if 2012 total circulation is similar to the 2010, ebook circulation would account for less than 4% of total circulation.
Looking at the percentage of library cardholders who have used the service will give a better indicator about usage and the actual size of the population of ebook users.
Percentage of total cardholders
In the period of January 2011 to June 2012, a total of 18,523 patrons registered for Overdrive among WPPL, Santa Monica, Omaha, and Robbins. Those libraries account for some 483,000 cardholders. Overdrive registrants represent 3.8% of cardholders.
|Table 4. Sample usage for lifetime of their Overdrive contracts|
|Library||Contract start time||% of card holders who have checked out at least one title from Overdrive during the contact period through June 2012|
June 2012 (Figure 5) was the month with the highest average percentage of cardholders that checked out a title via Overdrive out of the period from December 2011 to June 2012, the term for which there is data for all participating libraries. Unique users were determined by running a patron activity chart via Content Reserve.
When looking at the percentage of users who have used Overdrive over the lifetimes of the libraries’ contracts with the vendor, Figure 6 shows the individual percentages for the three libraries, whose data is included in this portion of the analysis, an average of 6.44%. Only data for three libraries is included in this chart.
One major factor of low ebook usage is marketing, as only 12% of library members were aware their libraries offered ebook services, according to aPew report. No report has attempted to capture the loss of existing Overdrive users, but this loss completes the picture of ebooks in libraries.
Given the tools available via Content Reserve, it is difficult to get data that indicates the loss of users. By looking at totals of registered users (registration presumably means entering their library card number) and comparing them to reports for unique users with checkouts, the difference between the two numbers is one indicator that can be determined. To be sure that the total of unique users did not include returning users, reports were run from the start of the libraries’ respective contracts with Overdrive up through June 2012.
For the three libraries included in this portion of the analysis, (WPPL, Robbins and Omaha in Figures 7 and 8), the differences between those two numbers indicated that 17%, 29%, and 30% of registrants, respectively, did not check out a title during the life time of the contract up through June 2012. This figure does not account for people who have checked out titles but then stopped using the service at some point, or those who decided not to use Overdrive before logging in. This percentage was used to create estimates of non-use, and the impact of new and returning users on circulation.
Table 5 shows the formulas used to create estimates of new active users, non-starters — patrons who registered with Overdrive but never checked out a book, and returning users, showing their impact throughout the time period of this study.
|Table 5: formulas to break users into categories|
|A: unique users with checkouts||Known value from Overdrive report|
|N: new patrons||New registrations. Known value from Overdrive report|
|V: new active users (patrons who joined the service and checked out a book in the same month)||N-O=V|
|R: returning users||A – V = R|
|O: non-starters (estimated)||N * P = O||Number of Overdrive registrants that did not check out an Overdrive title from the start of the library’s contract through June 2012.|
|P||(N-A)/N||Percentage of Overdrive registrants that did not check out an Overdrive title. Calculated for the entire term of the library’s Overdrive contract|
Checkouts by new users (Figure 7) follow the pattern for overall checkouts in Figure 4. As the pool of returning users gradually increases, so do their checkouts. In contrast to the hype of ebooks, adoption of ebooks by new users is stable and predictable, as is their checkout behavior. There is a separation between checkouts by returning and new users in Figure 8 from February to April, when the checkouts by new users drop off.
To evaluate the methodology for determining these estimates, the data for WPPL was analyzed for from the start of the library’s contract with Overdrive through June 2012. The estimated monthly totals of unique users with checkouts and non-starters were compared with known values from reports or from calculations. The differences between the estimated totals and the known totals were negligible as seen in Table 6.
|Table 6: Evaluating the estimates|
|Time period||unique users w/checkouts||new registrations||actual non-starters (total registrations – unique users with checkouts)||%of total registrants that did not check out a title|
|Known values from 6/1/10-5/31/12||
|estimated totals (totaling the estimated value for each month)||
|diff. in estimate from report total:||
These figures are estimates and should be considered as such. That said, the data indicate a significant level of usage loss. Using the reports available via Content Reserve, it is not possible to generate data to indicate patterns of usage for unique users in order to see at what point those who have checked out a title stop using the service, or at what intervals users check out titles. However, it can be shown that a significant number of patrons go as far as registering (which simply involves entering a library card number) but never check out a title.
The future of ebooks in libraries looks expensive. Driven by a desire to provide their members with as broad access to content as possible, libraries will be pressured to add multiple ebook vendors, and thereby assure that they will remain powerless against the decisions of publishers and vendors. That pressure comes in an economic environment when many libraries are dealing with decreased budgets:
“Twenty-three states reported cuts in state funding for public libraries from 2010–2011 to 2011–2012. For three years in a row, more than 40% of participating states have reported decreased public library funding” (American Library Association, p. 9)
The present ecosystem for ebooks in libraries does not represent a value for our members. While competitors to Overdrive have emerged, this competitive environment does not drive prices down, as prices are controlled not by vendors, but by publishers. Without the ability to own ebook content and migrate collections between vendors, then libraries cannot benefit from the arrival of new competitors in the marketplace. In the end, library ebook collections will remain fleeting, bound to vendors; and not only expensive to acquire, but also to sustain. An environment in which access to as broad a range of content as possible is secured for patrons, and affordable to libraries, will not emerge out of the current environment.
In consideration of the holiday boom in ereader and tablet sales, it is clear that readers’ acquisition of ereader or tablet devices drives usage of Overdrive services. Based on the major events noted in the sidebar above during the period of January 2011 to June 2012, libraries have progressively ended up with access to less content, as publishers either removed their content from the library marketplace, or imposed restrictive conditions on circulation and/or exorbitant prices that place additional financial strain on library ebook collections. While devices drive usage of library ebook services, the content that libraries have to offer is insufficient to sustain it.
American Library Association. (2012). The 2012 State of America’s Libraries. American Libraries, special issue. Retrieved online from: http://www.ala.org/news/sites/ala.org.news/files/content/StateofAmericasLibrariesReport2012Finalwithcover.pdf
American Library Association & Overdrive. (2012). Library eBook Survey hosted by OverDrive and American Library Association (ALA). Retrieved online from: http://overdriveblogs.com/library/files/2012/11/ALA_ODSurvey.pdf
Cape May County Public Library data. Data is from Public Libraries Survey, Fiscal Year 2010. Retrieved online from: https://harvester.census.gov/imls/search/library_list.asp?LibraryName=cape+may&LibraryID=&Address=&City=&State=NJ&Zip=&Distance=&County=&PhoneAreaCode=&Phone=&Librarytype=LS&Librarytype=CE&Librarytype=BR&ResultSetNumber=1&LibTypes=LS%2CCE%2CBR&StateSelectedIndex=31
Davis, Alex. (2013). Digital books outpacing print in some libraries. Shore News Today, Jan. 21, 2013. Retrieved online from: http://www.shorenewstoday.com/snt/news/index.php/2010-04-07-20-18-16/middle-township-news/33842-digital-books-outpacing-print-in-some-libraries.html
International Data Corp. (Mar. 13, 2012). Media Tablet Shipments Outpace Fourth Quarter Targets; Strong Demand for New iPad and Other Forthcoming Products Leads to Increase in 2012 Forecast, According to IDC. [Press Release]. IDC.com. Retrieved on line from: http://www.idc.com/getdoc.jsp?containerId=prUS23371312#.UTD28jB418E
International Data Corp. (Dec 5, 2012). IDC Raises Tablet Forecast for 2012 and Beyond As iOS Picks Up Steam, Android Gains Traction, and Windows Finally Enters the Market [Press Release]. IDC.com. Retrieved on line from: http://www.idc.com/getdoc.jsp?containerId=prUS23833612#.UTD30zB418E
Kelley, Michael. (Apr. 6, 2011). Kansas State Librarian Goes Eyeball to Eyeball with OverDrive in Contract Talks. LibraryJournal.com. Retrieved online from: http://www.libraryjournal.com/lj/home/890089-264/kansas_state_librarian_goes_eyeball.html.csp
Kelley, Michael. (Oct. 18, 2012). Random House Says Libraries Own Their Ebooks. LJ Insider. Retrieved online from: http://lj.libraryjournal.com/2012/10/opinion/random-house-says-libraries-own-their-ebooks-lj-insider/
Lovett, Michael. (Jan. 15, 2013).Digital Library As Hub of Discovery: 2012 OverDrive Data Review. Digital Library Blog. Retrieved online from: http://overdriveblogs.com/library/2013/01/15/digital-library-as-hub-of-discovery-2012-overdrive-data-review/
Miller, K., Swan, D., Craig, T., Dorinski, S., Freeman, M., Isaac, N., O’Shea, P., Schilling, P., Scotto, J., (2011). Public Libraries Survey: Fiscal Year 2009 (IMLS – 2011 – PLS – 02). Institute of Museum and Library Services. Washington, DC. Retrieved online from: https://harvester.census.gov/imls/pubs/pls/pub_detail.asp?id=140#
Overdrive. (April 16, 2012). Library Media Network Ebook Report. Retrieved online from: http://www.scribd.com/doc/94302507/Overdrive-Library-Media-Network-eBook-Report-Big-Data-1
Owen, Laura Hazard. (Feb. 10, 2012). Amazon’s Kindle Plays A Part In Penguin’s Library Decision. paidContent.org. Retrieved online from: http://paidcontent.org/2012/02/10/419-amazons-kindle-plays-a-part-in-penguins-library-decision
Owen, Laura Hazard. (Sept. 14, 2012). Hachette to raise ebook prices for libraries by 220%. paidContent.org. Retrieved online from: http://paidcontent.org/2012/09/14/hachette-to-raise-ebook-prices-for-libraries-by-220/
Price, Gary. (Jan. 24, 2013). Macmillan Announces Details of Library Lending Pilot. infodocket.com. Retrieved online from: http://www.infodocket.com/2013/01/24/macmillan-announces-details-of-library-lending-pilot/
Price, Gary. (Feb. 22, 2013). Macmillan Ebooks Pilot on OverDrive Begins In One Week. infodocket.com. Retrieved online from: http://www.infodocket.com/2013/02/22/macmillan-ebooks-pilot-on-overdrive-will-begin-in-one-week/
Publishers Weekly. (Oct. 29, 2012). Random House, Penguin Agree to Merge. Publishersweekly.com. Retrieved online from: http://www.publishersweekly.com/pw/by-topic/industry-news/industry-deals/article/54536-random-house-penguin-agree-to-merge.html
Swan, D. W., Grimes, J., Owens, T., Vese, Jr., R. D., Miller, K., Arroyo, J., Craig, T., Dorinski, S., Freeman, M., Isaac, N., O’Shea, P., Schilling, P. Scotto, J. (2013). Public Libraries Survey: Fiscal Year 2010 (IMLS-2013–PLS-01). Institute of Museum and Library Services. Washington, DC. Retrieved online from http://www.imls.gov/assets/1/AssetManager/PLS2010.pdf
Wright, Jazzy. (Nov. 15, 2012). OverDrive Survey Finds Library Patrons Buy Ebooks. District Dispatch. Retrieved online from: http://www.districtdispatch.org/2012/11/overdrive-survey-finds-library-patrons-buy-ebooks/
Zickuhr, Kathryn; Rainie, Lee; Purcell, Kristen; Madden, Mary; Brenner, Joanna. (2012). Libraries, patrons, and e-books. Pew Internet & American Life Project. Retrieved online from: http://libraries.pewinternet.org/2012/06/22/libraries-patrons-and-e-books/