Ebook Availability Revisited: A quantitative analysis of the 2012 ebook aggregator marketplace

Editor's Notes

• #4: This presentation repeats a study we presented at the 2008 Charleston Conference.Available: http://www.slideshare.net/john_mcdonald/charleston2008-ebook3
• #5: UPCC = University Press Content ConsortiumUPSO = University Press Scholarship Online
• #6: Details on the 2008 study.
• #9: The 2012 study used different methods to ask the same question.Four of the five libraries from the first study were includedWe started with the OCLC xISBN & WorkIDcrosswalks from the startWe used ebook knowledgebase data instead of vendor source files**One aggregator expressed concern that the KB based numbers looked quite low—We are validating our results by acquiring all of the source files direct from the aggregators to determine whether the 2012 marketplace (and match rate) is significantly underrepresented by the knowledgebase data.
• #11: C was correct; More 35-50%Using our updated protocol, 24% of 2006 (2008’s study) pub year print books were available electronically, while37% of 2011 (2012’s study) pub year print purchases were available Not status quo, but not a quantum leap either!The increase from 24 to 37% does represent more than a 50% increase…*(subject to validation using ebook aggregator source files)
• #12: We were interested in the perceived ‘quality’ of the available ebooks (some have speculated that ebook vendors are publishing lots of books but of lower quality).Availability does appear to vary by YBP select rating- E book versions were more likely to be available for research and specialized books than for Basic or Supplementary. There were no obvious differences between the two year samples in this comparison (data not shown). We speculated that a large portion of the books in the two Basic categories are University Press publications that are less likely to be in aggregator lists.
• #14: D – 100% was the correct answer (see next slide)Follow-up work based on aggregator-direct lists may show it to be even greater
• #15: 2008 inGreenvs 2012 in brown Percentages are portion of the total marketplace in that year (for green bars the proportion of the tallest green bar, and likewise with brown)Greatest market share increase was 22% ebrary, while Ebsco/Net Library decreased by 17%,perhaps not surprisingly given that their 77% share of the market in 2008 was based on their unique collection of older books rather than on that great a difference in acquisition of the current year’s monographic output. Average market share has increased slightly from 47 to 52%, but the range has dropped by 2/3rds (ie from 46 to 16%) If you were making a decision in the past on an aggregator based on the relative size of their collections, those differences are decreasing over time.So if each aggregator covers 50 % of the marketplace, it follows that two should cover most books, right?...
• #16: Unfortunately, that’s not the case--nearly 50% of aggregator-hosted ebooks are still offered by only ONE vendorAvailability across aggregators has increased significantly however: in 2008 only 1 in 25 books were available from all 4 whereas in 2012, 1 in 5 is available from all four
• #17: The vast majority of ebooks in aggregated products are those published in the past 15 years.with Ebsco/Net library showing clear dominance in terms of # of books offered each year for 1990-2004. The past 8 years (2005-2012) are worth looking at it more detail.
• #18: In 2008, MyILibrary (in blue) moved from the middle to the bottom in terms of coverage of the most recent years’ books, Starting in 2010, Ebrary (in green) outranked its competitors, and Ebsco has fallen in line with EBL over the past two years.
• #19: In 2011 Ebsco launched its own large academic subscription optionThe first competitor for ebrary’s Academic Complete subscription collection
• #20: From Aggregator provided source filesEbraryAC = Ebrary Academic CompleteEbscoES = Ebsco Ebook Subscription collection
• #21: Publication dates of subscribed collections mirror the full collectionsEbsco has many more books from the 80s & 90sWhen we take a closer look at the most recent decade…
• #22: Ebrary has a small percentage more from the first half of the decadeAnd things seem to be evening out for the most recent five yearsThere appear to be signs of a bit more differentiation for 2010 and 11 as the number of books in common is a lower proportion of the wholeAnd as expected there is about a year and a half lag before ebooks are added to these collections
• #23: The key point here is that there are a number of cases where particular publisher imprints are only listed for one aggregatorThough one might expect that this is the result of “exclusive” subscription deals, it is more likely caused by the newness of the Ebsco subscribed collection AND in some cases different publisher naming conventionsFor example ebrary does have books from Elsevier Science but uses the many subdivisions rather than grouping themThis same data is presented on the following two slides with a focus more on the publishers and less on the pattern
• #24: Keeping that caveat in mind, the blue and green bars here show cases were a publisher is listed in only one of the two collectionsAnd bar charts show the relative numbers for the publishers that appear in both collections
• #25: Keeping that caveat in mind, the blue and green bars here show cases were a publisher is listed in only one of the two collectionsAnd bar charts show the relative numbers for the publishers that appear in both collections
• #26: Based on LC class distribution to examine subject coverageOverall the distribution is quite similarEbraryACmay lean slightly toward science (see Medicine, Science and Technology) while EbscoES has more books in the humanities & social sciences
• #28: The data presented on this slide argue STRONGLY for subscribing to BOTH collections, given the differentiation presented in the previous slidesInstead of a 10 or 20 year period of subscription matching the purchase price, it would take 300(!) years of subscription costs to own the same contentFurthermore and perhaps equally important(!) books in the subscribed collections have unlimited simultaneous use, while purchased books from these two vendors are limited to a single simultaneous user (unless a premium is paid for each book)
• #29: 70-85% of the ebooks available in these University press collections are available from the major aggregators (albeit with more restrictive DRM)
• #30: But of course they represent only a tiny portion of what is available in the overall marketplaceJSTOR offerings are just becoming available as we speak
• #32: Each color slice shows the proportion of collection by decade or century*from hathitrust.org website [http://www.hathitrust.org/visualizations_dates]
• #33: Caveat– this only reflects relative size -- its not book to book matching as in prior comparisons…that’s a next step
• #34: Not a fan of 3d graphs, except in this case where the EBAM is effectively carpet on the floor that wouldn’t be visible if you looked at it end on.Although some currently question the value of our owned print collections, potential future subscription models from Hathi or Google (should copyright and royalty issues prove to be solvable) would definitely call into question the value of ownership of aggregator hosted ebooks that are subsumed within them