Introducing 

Prezi AI.

Your new presentation assistant.

Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.

Loading content…
Transcript

Buying by the Bucketful

Founded 1881

Research intensive (Russell Group)

Most disciplines:

  • Medicine
  • Biological Sciences
  • Veterinary Science
  • Physical Sciences
  • Engineering
  • Social Sciences & Law
  • Management School
  • Humanities

Fairly representative (?)

Since Oct 2006 we have purchased (or subscribed to) 19 eBook bundles

Mainly end of financial year purchases

A good way to spend end-of year funds for the benefit of the maximum number of staff and students

Better than e-journal backfiles?

Now budgets are tight...

  • Are these packages being used enough?
  • Are they giving us good enough value?
  • Can we afford them in the future?
  • Would PDA be better value?
  • Back to title selection?
  • Study the evidence...

Liverpool's experiences

with Springer...

Purchased all-subjects collections:

©2005-2008 in July 2008

©2009 in June 2009

©2010 in January 2010

©2011 in January 2011

Our STM depts choose to spend 95% of their library budget on journals.

But when eBooks are made available they account for 40% of usage on SpringerLink.

eBook purchases are a medium to long-term investment for the future, with no 'back history' of usage data.

We make journal renewal decisions based on the past year's use, mainly of previously-purchased articles published in previous years.

After 1 year of ownership about 45% of titles have been used

After 2 years of ownership about 65% of titles have been used

Experiences from Japan

Waseda University

  • eBooks are 44% of SpringerLink downloads

University of Tokyo

  • 40% of titles used in first year of ownership
  • 60% of titles after two years of ownership

Told you we were typical!

Are we seeing subjects starting to plateau out at 3 different levels?

Scope for prioritising which subjects to buy in the future?

Or will they all plateau out at about the same level and just take very different lengths of time to get there?

So if we only need 20% of the titles to get 80% of the downloads, the other 80% of the titles aren't needed and we just need to find a way to choose the right 20%...

Right?

WRONG!

Single eBook titles on

aggregator platforms

Chosen by academics, or

heavily-borrowed in print

3 titles (out of 465)

account for 21%

of full-text access

1 title (out of 190) accounts for 35% of full-text accesses

A third of titles have had 2 full-text accesses or fewer - in 33 months of ownership!

Is that any better than buying a package?

A slightly higher proportion of low use titles than a hand-picked collection.

But not bought at full price!

But with Patron Driven Acquisition all titles are used.

So that must be the best solution...

Right?

"models don't matter, only prices do"

really?

ebrary PDA model

Purchase triggered by:

  • 10 page turns in a title in a user session (not including front matter or index)
  • 10 minutes in a title in a user session
  • 1 x copy and paste from a title
  • 1 x print from a title

model this using COUNTER stats from packages

make assumptions:

  • Liverpool ebrary stats show average pages per session = 11

  • Assume each session is for a single book (sessions usually start at catalogue, so not unrealistic?)

  • 'Average of 11' means <11 pages read in half of sessions

  • Model by saying any book with 2 chapters downloaded or more in a Springer BR2 report would be a purchase under the ebrary PDA model

  • Conservative: says no books with a single chapter download would be purchased under the ebrary model

I'm cheating!

Our 2008 purchase included 2005-2007 at very little cost

But that's one of the benefits of buying packages

So let's take out 2005-2007 and adjust the 2008 payment down a little

3 package purchases in this period adds complexity

3 tranches of content exposed to the PDA model

Let's look at a single year's collection for clarity...

Duh... you did the modelling wrong then!

Iowa ebrary PDA pilot

  • Deposited $25k to start
  • 19k MARCs loaded (meant to load 100k)
  • Started Oct 1
  • By Nov 30 spent $28k on 262 titles
  • Weekly spend increasing
  • STOP!
  • Bought Wiley, Springer, Elsevier packages
  • Slashed the titles exposed to PDA
  • Carry on...

other PDA models have a higher purchase threshold, but:

rental costs for not-yet-purchased titles

purchase titles at > list price because of prior rental

For price of package, could only buy titles with 7 or more chapter downloads at list price

What PDA model is going to better that?

Evidence-based purchasing

  • Pay a modest sum up-front
  • A year's access to everything
  • At the end decide what to retain access to
  • Up to the price you've already paid
  • Unless you want to pay more

Foolproof?

We've already done it!

  • Started July 2007 with a year's access to everything
  • After 1 year purchased ~580 titles with 7,150 downloads
  • In Year 2, 77% were used, 4,686 downloads
  • In Year 3, 73% were used, 5,848 downloads
  • 61% were used in all 3 years
  • What about the titles we didn't buy?
  • How well would they have been used in years 2 and 3?

But a useful evidence base to agree on fair pricing for a package?

Other things to think about...

Crisis in academic monograph publishing

  • Print run 3,000 in 1980, 350 now
  • Wouldn't OA be great?
  • But how to fund it?
  • Who's awash with money? Libraries! (??)
  • International Library Coalition for Open Access Books (ILCOAb)

Conclusions

My ideal (e) books portfolio:

  • Packages where they give best value

  • Full-text eBook databases to give a cheap critical mass of 'stuff'

  • Single title selections for core texts

  • PDA to fill the gaps, not to form the foundations

Much like our (e) journals portfolio in fact

Implications for libraries:

  • Need to centralise book budgets

  • Rapid movement to e-only book acquisition

  • A hard sell to (some) faculty

  • Implications for space, staffing, services

  • Spend more on e-content, less on servicing dead trees

Implications for publishers:

  • Discounts of 10%, 20% or 25% (or nothing!) off total list price do not make an attractive package proposition - PDA would be cheaper.

  • Need to see a 50% discount for an offer to be worth considering

  • At a discount of 70-75% its probably a no-brainer (if you publish relevant content)

  • Better to sell packages to lots of libraries at low cost, than to a few libraries at high cost, or a few books to many libraries at full price?

  • Boost sales of eBooks, give libraries cuts to journals expenditure by bundling 'content' together and dividing the price differently

War is declared at the ASA Conference

Charge much less, sell much more

Mark Ware, ASA Conference 2011

"can't pay"

"Tough. Get more efficient (like publishers already have)"

Phil Sykes,

University of Liverpool

"won't pay"

David Hoole,

NPG

David Prosser,

RLUK

Frances Pinter,

TOC 2010

Learn more about creating dynamic, engaging presentations with Prezi