I forget where I read the foregoing--some other library blog, perhaps--but I've been thinking about it a lot this morning, because I think it's true.
In my pre-library life, I worked for Amazon.com. This was in the company's very early days: a single warehouse in Seattle's SoDo district, a customer service department that sometimes went to help pack book orders (and only book orders, in those days) when things were slow, Jeff Bezos's famous laugh echoing down the stairs from his crow's nest of an office. At the time, people didn't know what to make of a company that didn't even have a storefront, let alone any stock.
And here's how they did it: they relied extensively on distributors, especially Ingram, which has an enormous warehouse along I-5 in central Oregon, to deliver books every morning based on orders that had been placed over the previous few days. In other words, anything that was in Amazon.com's warehouse was there because there was already an order for it. This was quite a change from traditional bookselling, and not just because orders were only placed online (people were surprised that there was no print catalog, either); brick-and-mortar retailers place orders based on what they think will sell, not what has actually sold.
So that's cool. But in order for it to work, there needed to be a distribution network that could deliver books to Amazon.com's warehouse quickly, so that from the customer's perspective, Amazon.com's service was in turn rapid and efficient.
What does this mean for libraries?
There's been a lot of discussion in recent years about resource sharing, moving beyond the basic ILL scheme to consortial agreements, alliance arrangements to speed up interlibrary lending, and so on. As library resources (particularly journals) get more and more expensive, it makes more and more sense to not buy something if the library down the road, with which you have an arrangement, has bought it.
But, that does require that other library to own, or have access to, that resource. Without that, you can't do your just-in-time delivery.
My library's ILL service is the fastest I've ever encountered. I've placed an article request in the morning and received it by the afternoon. Book delivery is necessarily slower but I still typically get my requests inside of a week. Our own collection is relatively small and curriculum-oriented, which means that, like most other faculty, when I'm doing my own research I make extensive use of ILL.
But in order for it to be as fast as it is, we need the access that we have: to an extensive regional network of academic and public libraries which includes a major Research I university with an award-winning library system, as well as one of the most highly regarded public library systems in the country.
It seems to me, then, that what just-in-time really does is push the just-in-case further back along the supply chain. There are advantages to this from the supplier perspective; publishers have just as much trouble figuring out how many books to print as booksellers have figuring out how many books to buy. And as long as there are libraries with a mandate to preserve as well as to provide access, as that Research I up the road has, then libraries like mine, where access is the main guiding principle, can continue to provide quality service even as the resources get more and more expensive.