One really good indicator of value, or ROI, is cost per use. If you have a journal or database that costs $10,000 per year, and it gets used 100,000 times in a year, then your cost per use is only $0.10. If you journal costs $10,000 per year, and it gets used only 1,000 times, then your cost per use is $10.

Whether $10 per use is too high or just fine is up to you to determine. It might seem reasonable, because you would spend $40 to purchase the article, or maybe $20 to get the article through interlibrary loan. Or you may be part of a good consortium, and ILL is efficient and free, so $10 is too much to pay.

For a library, the views of potential users community is prior who can absolutely focus on objective of the institute. After that there are some other valuable indicators which are practiced core journals in a discipline, citations, download count, publication history of journal as well as its publisher etc.

Hi Scott,

I agree with you, this is the first one. It is extensively used by libraries, and we do it. But it's what I wil call a "carpet bombing indicator". I think we need to go further and be more selective and relate indicators not only to simple macroeconomic rules but to introduce some quality or microeconomic rules.

Consider two journals with the same cost per article. All the considerations of interlabrary loan and consortium membership being equal. I think we need to consider some more parameters like for instance the size of the research communities interested in these two journals. If the price of the subscription is the same, if the number of dowloaded articles is the same, if journal J1 diserves a community about 10 times smaller than the community diserved by tje journal J2, then J1 is about 10 times more read than J2.

Same type of consideration maybe taken into accounts for readers status (students ? Undergraduates ? Graduates ? Researchers ? ...). The subscription price if most often based on the number ot potential readers. And we have interest in moving to a number of potential readers as close as possible to the number of real readers to get lower prices, as far as the economic model run by the editor allows that possibility. Precise knowledge of readers communities will contribute to lower prices.

This should conduct to more taylored scientific journal subscribtion policies, more selective towards researchers needs in institutions and less blind cuts due to budget pressure. The disccusion is stil open. No ?

Hi Jn,

We are on the same wavelength. I was wondering if this only qualitative feelings or if these indications" have already been quantified ?

Hi Didier,

Many professionals have conducted study on journals collections for their library. May be Impact Factor devised by ISI, Philadelphia has some drawback for quantitative study but mostly the list of high impact journals coincidence with views of the researchers in regard to high impact journals of the globe.

When money is not a constraint, the indicators favor more the scientific community to subscribe to journals are: visibility and relevancy of the titles to be acquired, this is based on the Law of Bradford

Thanks all,

I agree on core journals. there are no doubt. The point was abouut where to cutt and how when budget constraints increase. How to cut not only from the econimic point of view (cost per use) but favoring science.

We have moved increasingly from a custodial 'just in case' policy to a rapid response 'just in time' policy, requesting articles on demand from a European consortium. This allows us to use our budget flexibly to the best advantage. Another option when sourcing individual articles is to pre-purchase credits with publishers.

Asking users directly what they need from your service can lead to confusion (some want one title, some another). Users are often shocked at the cost of journal subscriptions so tell them the truth if you can't afford something.

Check out journal and article impact factors as a mark of quality.

Thank you Vaughan,

Did you evaluate the savings with the "just in time option" versus "the just in case one" ?

It wasn't really a case of saving but deploying the budget more effectively. Instead of spending €15.000 to sit on a shelf (or a server) and wait to be used, we spend it on responding to genuine needs linked to current, on-going research projects.

Keeping a watch on journal usage reports for our on-line titles also gives an indication of the true value of a journal to the organisation. Users will often say you *must* subscribe to a title. If you can show them it is used less than once a month (for example) you have a case for dropping the subscription.

Another possibility is to use the Eigenfactor Cost Effectiveness metric at http://www.eigenfactor.org/costeffectiveness.php

Cost Effectiveness (CE) is created by dividing the yearly subscription price by the Eigenfactor Score (a measure of journal influence based on citations and journal size). It ranks journals by a number of categories in science and social science, through some of the subjects are pretty broad (such as medicine, education)

I also agree it is very important to keep an eye on journal usage at your institution. If a journal is not being used it probably does not make sense to keep a subscription. Also, in the past Librarian may have been more reluctant to stop journal subscriptions because there would be a gap in the print if it needed to be restarted. Online journals provide more flexibility in this regard.

I would also look at where your faculty publishes, and the programs at your university, especially new ones that might be growing in importance.