Choosing how to deliver ebooks to our users is no easy task. There are many different publisher and vendor platforms, and there are just as many access models. A librarian that wants to build a broad ebook collection must consider multiple platforms and models. This choice inevitably involves a decision about access to a wide range of content through DDA or subscription packages versus the security of ownership. It is clear that DDA and subscriptions allow far more titles to be made available to a library’s users. This study explores the value of the various access models by measuring usage, cost, and holdings at both the local and consortial level.

1. Making Value Judgments:
eBook Pricing for Access and
Ownership
Electronic Resources & Libraries
Austin
February 23, 2015
Michael Levine-Clark
University of Denver
Jason Price
SCELC
Maria Savova
Claremont Colleges

2. Elements of Value
• Long-Term Stability
• Ease of Access (DRM)
• Staff Time
• Choice/Flexibility
• Cost
• Usage

3. Access Models: Balancing Rights, Cost,
Long-Term Access
Temporary  Access  Perpetual
LowCurationHigh
LowCostPerUseHigh
Heavy  DRM  Light
Aggregator
Publisher

4. Access Models: Balancing Rights, Cost,
Long-Term Access
Temporary  Access  Perpetual
LowCurationHigh
Title-by-Title
EBS
Package
DDA without STL
DDA with STL
Subscription
LowCostPerUseHigh
Heavy  DRM  Light
Title-by-Title:
• Heavily customizable
• Staff time
• High cost per unit
• No guarantees of
usage
• Perpetual Access
• If on publisher platform,
light DRM

5. Access Models: Balancing Rights, Cost,
Long-Term Access
Temporary  Access  Perpetual
LowCurationHigh
Title-by-Title
EBS
Package
DDA without STL
DDA with STL
Subscription
LowCostPerUseHigh
Heavy  DRM  Light
Subscriptions (especially on
an aggregator platform)
• Little to no customization
• Content can be removed
at any time
• Heavier DRM
• Lots of content for
relatively low cost
• Not much staff time
• Helps build critical mass

6. Access Models: Balancing Rights, Cost,
Long-Term Access
Temporary  Access  Perpetual
LowCurationHigh
Title-by-Title
EBS
Package
DDA without STL
DDA with STL
Subscription
LowCostPerUseHigh
Heavy  DRM  Light

7. Subscription Packages
• ACLS HEBS
• Books 24x7
• ebrary
• EBSCO
• Knovel
• Safari
DDA
• EBL
• ebrary
• JSTOR
EBS
• CRC
• Palgrave
~$510,000
Perpetual Access
• Duke
• ebrary
• EBSCO
• OUP
• Springer
19%
49%
25%
7%

8. Subscription Packages
• ACLS HEBS
• Books 24x7
• ebrary
• EBSCO
• Knovel
• Safari
DDA
• EBL
• ebrary
• JSTOR
Perpetual Access
• Duke
• ebrary
• EBSCO
• OUP
• Springer
EBS
• CRC
• Palgrave
~575,000 titles
46%
36%
13%
5%

9. Cost vs. Availability

10. Cost vs. Availability

11. Assessing Value: Long-Term Stability
Temporary
Access
Perpetual
Access

12. Long-Term
Stability
Title-by-Title Purchasing | Publisher-Hosted Packages
All content is permanently licensed by the library

13. Long-Term
Stability
Evidence-Based Selection
All titles available during selection period,
with perpetual access to selected titles.

14. Long-Term
Stability
DDA
• Temporary access to
consideration pool,
and titles can be
removed by
aggregator.
• Perpetual access to
purchased titles

15. Long-Term
Stability
Subscription
• Temporary access
• Titles can be removed

17. Ease of Access
• Digital Rights Management (DRM)
– Restrictions on access (single vs multiple vs
unlimited users)
– Restrictions on actions
• Printing
• Downloading
• Copying

18. Publisher-Hosted Content
Typically little or no DRM
Packages, Title-by-Title,
EDS

19. Aggregator
Typically more DRM
User limits, limits on
functionality

20. Staff Time
• How much time does it take staff to manage
selection, deselection, profiling?

21. Staff Time
Subscription
No staff intervention beyond annual renewal, periodic record loads

22. Staff Time
Publisher-Hosted Packages
One transaction to get multiple titles;
depending on publisher, may need to select multiple packages

23. Staff Time
EBS
Select publisher package
Annual selection of titles

24. Staff Time
DDA
• Initial profiling,
periodic adjustments
• More staff time if
mediated
• Record loads

25. Staff Time
Title-by-Title Selection
• Regular selection
of titles

26. Choice/Flexibility
• To what extent do librarians get to select titles
for purchase or insertion into consideration
pool?

27. Choice/Flexibil
ity
Title-by-Title Selection
Full choice of titles (but limited to whatever is available from that vendor)

28. Choice/Flexibil
ity
DDA
Creation of profile, ability to manually add/remove titles
Limited by title availability from vendor/DDA eligibility

29. Choice/Flexibil
ity
EBS
Limited ability to choose initial package
Full control of titles purchased

30. Choice/Flexibil
ity
Publisher-Hosted
Packages
• Must take entire
package
• Once purchased,
titles remain in
collection

31. Choice/Flexibil
ity
Subscription Package
• No choice – package is
preselected
• Titles come and go

32. COST & USE

33. Cost
• Cost alone doesn’t tell much
• Cost per title available is more meaningful

34. Cost per Title
Available
Subscription
Cheapest cost per title available. Tradeoff is no long-term access

35. Cost per Title
Available
DDA
Only titles with use paid for, but large pool of titles available.
Browse, STL programs cheaper than auto purchase.

36. Cost per Title
Available
EBS
Upfront cost, select titles based on use,
but large pool available

37. Cost per Title
Available
Publisher Package
Discount to purchase in
bulk, but more expensive
per title than DDA, EBS, or
Subscription.

38. Cost per Title
Available
Title-by-Title
Purchasing
• Generally list price for
each title

40. Use
• Use alone doesn’t tell much
• Cost per use more meaningful
• Cost per title used has some meaning as well

41. A Use is not a Use is not a Use
• COUNTER BR2 – Successful Section Requests
– Not always defined
– Doesn’t say anything about the type of use

42. Other Data Problems
• Non-use often excluded from reports
– Difficult to retroactively determine full title lists
– Non-use helps tell the story of use
• COUNTER reports often conflate business
models (subscription, perpetual access, DDA)
• List price for titles generally not included in
reports

43. Time Series Data is Difficult
• How do you account for years in which you
pay for backfiles?
– Would be easier if you could track usage
separately
• Need to track owned/not-yet owned content
separately
– EBS
– DDA

44. Cost Per Use (CPU)
• Measures cost of each use = actual use
• Does not account for different types of use
– Longer vs shorter use
– Different measures by different vendors
• Is institution specific
– A school that can afford to spend a lot of money
(or is forced to because of historical spend) but
has a relatively small user base (or large collection
relative to user base) may have a high CPU
compared to a school with opposite
characteristics.

46. Cumulative cost per use
(package level)
• One time book package purchases gain usage
over time  cost per use (CPU) decreases
• Annualized Package purchase CPU cannot be
compared to subscription CPU
• We need a cumulative definition of CPU

47. Cost Per Use –
Annual vs Cumulative (Platform Level)
Publisher hosted ebook
collection
Year 1 Year 2 Year 3
Total Books Owned 5000 6000 7000
Books Purchased 5000 1000 1000
Cost $50,000 $30,000 $30,000
Use 1000 1300* 1600
Annual Cost per use $50 $23 $19

48. Cost Per Use –
Annual vs Cumulative (Platform Level)
Publisher hosted ebook
collection
Year 1 Year 2 Year 3 Year 4
Total Books Owned 5000 6000 7000 7000
Books Purchased 5000 1000 1000 0
Cost $50,000 $30,000 $30,000 $500
Use 1000 1300* 1600 1000
Annual Cost per use $50 $23 $19 $0.50

49. Cost Per Use –
Annual vs Cumulative (Platform Level)
Publisher hosted ebook
collection
Year 1 Year 2 Year 3 Year 4
Total Books Owned 5000 6000 7000 7000
Books Purchased 5000 1000 1000 0
Cost $50,000 $30,000 $30,000 $500
Use 1000 1300* 1600 1000
Annual Cost per use $50 $23 $19 $0.50
Cumulative cost $50,000 $80,000 $110,000 $110,500
Cumulative use 1000 2300 3900 4900
Cumulative Cost per use $50 $35 $28 $22

50. Cost Per Use –
Annual vs Cumulative (Platform Level)
Publisher hosted ebook
collection
Year 1 Year 2 Year 3 Year 4
Total Books Owned 5000 6000 7000 7000
Books Purchased 5000 1000 1000 0
Cost $50,000 $30,000 $30,000 $500
Use 1000 1300* 1600 1000
Annual Cost per use $50 $23 $19 $0.50
Cumulative cost $50,000 $80,000 $110,000 $110,500
Cumulative use 1000 2300 3900 4900
Cumulative Cost per use $50 $35 $28 $22

51. Cost Per Use –
Annual vs Cumulative (Platform Level)
Publisher hosted ebook
collection
Year 1 Year 2 Year 3 Year 4
Total Books Owned 5000 6000 7000 7000
Books Purchased 5000 1000 1000 0
Cost $50,000 $30,000 $30,000 $500
Use 1000 1300* 1600* 1000
Annual Cost per use $50 $23 $19 $0.50
Cumulative cost $50,000 $80,000 $110,000 $110,500
Cumulative use 1000 2300 3900 4900
Cumulative Cost per use $50 $35 $28 $22

52. Cost Per Use –
Annual vs Cumulative (Platform Level)
Publisher hosted ebook
collection
Year 1 Year 2 Year 3 Year 4
Total Books Owned 5000 6000 7000 7000
Books Purchased 5000 1000 1000 0
Cost $50,000 $30,000 $30,000 $500
Use 1000 1300* 1600* 1000
Annual Cost per use $50 $23 $19 $0.50
Cumulative cost $50,000 $80,000 $110,000 $110,500
Cumulative use 1000 2300 3900 4900
Cumulative Cost per use $50 $35 $28 $22

53. Cost Per Use –
Annual vs Cumulative (Platform Level)
Publisher hosted ebook
collection
Year 1 Year 2 Year 3 Year 4
Total Books Owned 5000 6000 7000 7000
Books Purchased 5000 1000 1000 0
Cost $50,000 $30,000 $30,000 $500
Use 1000 1300* 1600 1000
Annual Cost per use $50 $23 $19 $0.50
Cumulative cost $50,000 $80,000 $110,000 $110,500
Cumulative use 1000 2300 3900 4900
Cumulative Cost per use $50 $35 $28 $22

54. Consortium Cumulative CPU – Publisher Package Purchase 1
$577.71
$193.81
$133.52
$0.97
$0.78 $0.71
$25.52
$16.72
$12.72
$0
$1
$10
$100
$1,000
yr1 cum yr2 cum yr3 cum
CumulativeCostperUse(LogScale)

55. $5.08
$1.33
$0.38
$0.00
$1.00
$2.00
$3.00
$4.00
$5.00
$6.00
FiveYearCumulativecostperuse
Institution
Consortia subscribed collection
Five year cumulative cost per use

56. Cumulative CPU comparison
$54.36
$8.91
$4.17
$10.94
$8.05
$6.60
$5.59
$4.84$4.83
$2.79
$1.67
$0.69
$0.93
$0.77 $0.74 $0.78
$0.10
$1.00
$10.00
$100.00
After 1 year After 2 years After 3 years After 4 years After 5 years
CumulativeCostPerUse(Logscale)
Publisher package purchase 2
Individual e-book purchases
Publisher package purchase 1
Subscription package

57. CPU levels
• Title
• Package
• Platform
• Strategy

58. Challenges in Evaluating value going
forward
• Annual cost per use not sufficient
• Comparing annual purchase models with
subscription models – what’s valid and what’s
practical
• What is the value of access (to books not [yet] used?)
• What is the value of ownership to titles that
might otherwise be maintained by subscription?

59. Weighing the value of ownership, availability & use
3.5
$28.76
$34.57
12.2
$8.19
$4.23
1.0
$5.23
$24.85
0.4 $0.90
$7.42
$0.17
$1.23
$0.73 $0.92
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
Titles owned per $100 Cost Per Title available Cost per use
CostORNumber
Publisher A Package
Publisher B Package
Publisher C EBS
DDA with STL
Specialized Aggregator Subscription
Major Aggregator Subscription

60. Questions?