NFAISOnline Usage Statistics: Current Statusand Future Directions 27th of October 2006 Philadelphia

1
NFAISOnline Usage StatisticsCurrent Status and
Future Directions27th of October
2006Philadelphia

• Using Usage Statistics Usefully

Presented by Hanneke Steuten Elsevier
ScienceDirect Scopus Usage Research
2
Agenda

• The value history of usage reporting at
  Elsevier
• How is it being used?
• Wrap-up

3
The Value of Usage Data
4
Value of Usage Data

How can you determine if the products you buy are
valuable?
Give your customer the means to measure the value
of your product
i.e. usage reports!
5
Library Usage Reporting (History)

• 1999 Paper reports mailed/faxed to customers
• Dec 2000 5 usage reports were created for
  customers as an attachment sent by email
• March 2003 Major new release enabling librarian
  log-on and remote downloading
• Nov 2003 COUNTER Release 1 of the CoP for
  ScienceDirect
• Sep 2005 COUNTER compliance Release 1 of the CoP
  for Scopus
• Nov 2005 New User interface, integration with
  Librarian Admin tool
• Mar 2006 COUNTER Compliance release 2 of the CoP
  for ScienceDirect
• Working on SUSHI Elsevier has representative on
  SUSHI Committee

.Customers run over 150.000 reports on a yearly
basis!
6
Value for Elsevier
Direct feedback on what is valued
And what isnt
7
Internal Usage Reporting (History)

• 2000 high level reports showing general SD usage
  and downloads per journal
• 2002
• Internal awareness campaigns about the value and
  importance of usage
• Usage became a key performance indicator for
  publishing and sales regions
• 2003 new reporting system - users within sales
  and publishing run there own reports

Internal users run ca 55.000 reports on a yearly
basis
8
How is it being used?

• Many internal customers use usage data in their
  decision making processes
• IT Operations
• Publishing
• Product Development
• Sales Marketing

9
Publishing

• Usage is key indicator in Journal (Portfolio)
  Management
• How does the usage of my journals compare with
  others?
• What is the geographical distribution of my
  readership?
• What are the most used articles?
• How is current content used versus older
  material?
• What are the differences between subject areas?
• Usage data is reported back to editorial boards
  and societies

10
Publishing Citations versus Usage

• Study which analysed citation behaviour versus
  journal usage
• behaviour
• ISI impact factor article citations
• The existing quantitative indicator, author
  behaviour, formal, data gt 2 years old,
  comprehensive
• Journal usage data
• New, Reader author behaviour, informal, very
  up-to-date
• Conclusions
• In some fields authors and readers comprise more
  or less of the same group
• In other fields there is a large additional group
  who read only

11
Publishing Article usage versus citations
Trends Journals
Percentage Share over Articles Published
Earth Sciences basic
Nuclear High Energy Physics
Business Managemenbt
Fundamental Life Sciences
Ergonomics, Computer Engineering
Ecology Environmental
Nuclear Materials, Metals
Social Sciences
Chemical Engineering
Citations per subject area
Article downloads per subject area
12
ScienceDirect Scopus Product Management

• Analyze behavior of users and optimize products
  and services
• What features on the platform are used most?
• How many regular users do we have?
• How do user navigate through the product?
• Where are users coming from?
• What are the exit pages?
• What drives usage growth?
• What trends changing user behavior do we see?

13
In 1998 we knew next to nothing
Product Management SD Redesign

• Except
• Customers wanted
• a stable robust platform
• a scalable system
• high performance reliability
• No blue prints to copy, so we copied the current
  user behaviour

14
The navigation mimicked the researchers
behaviour at the time..
Go to the library
Go to ScienceDirect
15
Find the Right Subject Section
16
Select the right journal
Etc., etc.,
17
Product Management SD Redesign

• Path analysis shows that users have short user
  sessions
• 70 of all sessions consist of up to 5 user
  actions
• Existing mainly of basic functionality actions
• User Behaviour is changing
• Significant increase from linking from library
  homepages and AI databases that take the user
  directly to the full text article visa-vis
  browsing.
• Session became shorter between 2001 2004
• Number of articles retrieved has increased
  relative to the time of the average user session

• Make the user as efficient as he can be, make the
  product as easy
• as possible

18
Product Management SD Redesign

• Vision Redesign
• Accelerate time to results (less clicks)
• Enable improved customisation (stickiness)
• Focus on task at hand
• Empower users to save time, be more efficient and
  productive

19
Sales Marketing

• Keeping eye on customer usage levels and
  customising content offerings
• What is a customers Cost per Article? How does
  its usage level compare to similar institutes?
• How is a customers usage divided over subscribed
  lists versus e.g. collections?
• Which backfile offering would be interesting for
  this customer?
• Measuring effectiveness of campaigns, using usage
  data in campaigns
• What is the conversion rate of my online
  campaign?
• Using Top 25 Article alerts as a marketing tool

20
Sales Marketing Backfiles

• What backfile package would be a good investment
  for the customer?
• Assumption about usage 85 v 15 (Tenopir
  King)
• Strong variation across subject areas
  (Mathematics in Delft 43 v 57) life sciences
  somewhat lower (90 v 10) physics, mathematics,
  engineering somewhat higher (80 v 20)
• Actual demand viewings of free abstracts
  published before 1995 clustered by BF package,
  irrespective of whether BF has been purchased

21
Sales Marketing Backfiles
Small Technical University
Eng. Tech. ROI 2.9
Nurs. Health ROI 91.3
22
Sales Marketing Top 25
23
So..

• Usage as a key metric is now well integrated
  within (most) departments and their decision
  making processes internally
• Yet, future improvements to be made within e.g.
  the marketing campaign tracking area
• Usage is well tracked and measured, but how easy
  is it to influence?
• Usage growth is the result of a complex set of
  generic industry factors, journal characteristics
  and customer e-library facilities.
• Currently the key metric is article downloads,
  but can we get to a more weighted measure?
• What is good usage? Can we make a relative usage
  metric? Can we calculate a so called Usage
  Factor?

24
Thank you!