Notes from a presentation given by Helen Henderson (Information Power) at the Ingenta US Publisher Forum in 2006.
EDIT: slides now available -- download .pps
What are Institutional Identifiers? They are not new, there are many different types:
- delivery code
- lending code
- SAN, MARC Orgnzation codes, GLN, ISIL
- Tax identifiers
- Company numbers
No true unique identifiers for institutions.
What are they not?
- A passport
- Not a single number (multiples, granularity).
Institutions with multiple parts, and multiple owners
- Intelligent (too complex)
- URLs (changing all the time)
What metadata is needed?
Unique id, location, category, tier, size, URL, credentials. Data requirements vary through the supply chain, i.e. different facets for different uses.
Why do we need them?
Licensing, marketing, customer analysis, authorization, authentication, optimising support of the journal supply chain.
- Big deals: consortia, opt-in, and opt-out, global
- Changes in groupings.
- Assess Market penetration
- Identify gaps
- Institution overview (authors, editors, referees, customers).
E.g. assess contribution of institution to your content.
Sources of information
- "Customer" databases
- Accounting system
- Fulfilment System
- Sales management system
- ...etc, etc
Issues with accuracy of information through the supply chain. Intermediaries have different levels of detail, so pass on different bits of information. Often difficult to identify the actual subscribing institution. It's also useful to get information about non-subscribers, i.e. to whom should I market?
Customer analysis: group, e.g. by country, type, region, etc.
Market analysis: which of the top 50 subscribe, and what is our total revenue? Degree of market overlap with potential business partners?
- Supply chain increasingly complex
- Publishers have more direct publisher contact
- Technology allows easier research into institutions (couldn't have done this pre-Google)
- Increase in bulk negotiations. Example of OUP losing money on a site licence, because they didn't have a good overview of who was included
- Mixed institutions
- New authentication mechanisms
Conducting industry wide pilot. Mapped out a number of different "transactions" through the supply chain:
- Customer -> Agent
- Agent -> Publisher
- Publisher -> Distributor
- Distributor -> Hosting Service
- Hosting Service -> User
Participants in current pilot: BL, HighWire, Ringgold, Swets, UK Libraries. Project website is at www.journalsupplychain.org. Progress reports on each of the 9 work packages have recently been added. Intending to produce an evaluation report at the end of June.
Biggest issue is the metadata: what to attach for each transaction?
Benefits for users include improved activation. Helen noted that Ingenta has been quite aggressive at allowing agents to activate on behalf of customers (as far as she's aware we're the only company doing this).
Easy access to archives
Could also offer registry facilities (central IP registration). Although unlikely that libraries will want it, if there was a central trusted registry, then that would be of benefit to them.
Helen recently wrote a report on institutional identifiers for the Charleston Advisor.
Q: Others are doing this also?
A: Atypon are largely creating a database of IP addresses, there's no metadata. Ringgold are creating a database of institutions. We may fold in their work. Intention is that database will be in public domain. Would need sustainable business model, and are considering what that might be.
Q: How will you get buy-in from libraries
A: To certain extent, no need for buy in, they just need to use it. They should be able to see the direct benefits to them. Key benefit would be to see and immediately activate their whole institutional holdings in one go, not title by title. In terms of supplying their id with their order: that would be nice. Best thing to do to get buy-in is to put up something wrong and get people to correct it!