Usage statistics

Back in the old days Librarians yearned for a way to measure which journals in their collection were being used and how often. They would resort to putting sign-up slips on volumes, to be initialled every time a reader consulted that volume, or to paying library assistants to lurk in the background and spy on readers to see which titles were used. "If only", they said, "we had a reliable method".

And then came electronic journals, followed closely by electronic journal usage statistics. Suddenly we are drowning in data that we struggle to interpret, though combining it with cost data can be instructive. What's worse, the publishers also have this data and they are looking at it.

Science switched a few years ago to charging us according to the level of usage made. In the last renewal of Nature and other Nature Publishing Group (NPG) titles, the publisher presented figures to show that the cost per downloaded Nature article at NIMR was very cheap compared with the majority of other customers (and at LMB cheaper still - they seem to be heavier users). Hence, they told us, we should not complain about the price increases. Other publishers have also been looking at usage figures to see whether they can be a basis for charging, and we may see new business models based on usage emerge in the next few years.

I learned one interesting snippet from the Company of Biologists, publishers of the journal Development. Apparently use of this title at NIMR is about half that of the usage across the whole of Harvard Medical School.