Testing some examples I have of Umlaut’s links to Amazon prior to a presentation I’m giving tomorrow, I’ve noticed that sometime very recently, Amazon has significantly changed the interface and access controls to their ‘reader’ interface–the functions to provide ‘look inside’ and ‘search inside’ features.
These changes are mostly not for the better from my perspective.
The search results for ‘search inside’ used to show up in the main content area, in a very readable format, wtih excerpts. Now they show up in a tiny little box on the lower right. I used to prefer Amazon’s search-inside results to Googles, no debate—now Google’s are much better.
This is actually the big and interesting aspect. Amazon always required you to log in to view more than a few pages of content. Now, it appears to me that the amount of content you can view without logging in is much reduced. The way they’ve restructured what content shows up where and when, the interface is now much more annoying and less useful if you haven’t logged in. But much more significantly and worse than that, just having an Amazon account isn’t enough.
Please Sign InYou must be signed in and have made a prior Amazon purchase to view this page. Please sign in to your Amazon account.
Do I need to be signed in to preview books?No, but some features and pages are available only if you are logged into your Amazon account and have made a prior purchase. For those customers who choose not to log in or who have not made their first purchase with us, we still provide a helpful set of preview pages within the book.Does my prior purchase need to be a book?No. Any item sold by Amazon.com will do. (Digital items and items purchased from another seller through the Amazon site do not qualify at this time.) You must have a shipped order in your order history. If you need assistance changing your e-mail address or recovering your password on your old account, please contact Customer Service.
I can only imagine that Amazon had enough internal pages using these URL formats that they couldn’t conveniently change, that they had to keep those old URL formats working.
The screen-scraping I was doing to discover if an Amazon look-inside or search-inside is available is mostly working, but seems to have some issues. I think that sometimes Amazon does have a search-inside, but my routines aren’t succesfully finding it. Haven’t had time to debug it entirely, but I think it might have to do with me looking for the phrase ‘search inside this book’, which Amazon now sometimes changes to ‘search inside another edition of this book’ (which is actually helpful additional info for the user, and interesting that Amazon is working more on the multiple-edition issues we think about a lot).
Anyone want to wildly speculate on what motivated these changes? I have some ideas, but am not really sure. I kind of suspect that some of them may have been directed specifically to interfere with the kinds of ways I was trying to use Amazon, providing amazon look/search inside functionality to my users (who may have no intention of buying from Amazon) somewhat independently of the Amazon website. But maybe that’s just my own paranoia and delusions of grandeur.
So, the interface I am directing my users to for searching/looking inside a book is still pretty much consistent with what you get from Amazon itself. Some user interviews we did indicated that many of our users find (or found, the previous version) Amazon’s look/search inside very helpful, and often used it together with or instead of our library catalog. So I think it’s still a net win to pre-check availability of and link to these functions in my interfaces.
But those interfaces, whether linked to from my interfaces or discovered solely through Amazon’s entry points, have, to my eyes, become a lot less useful, less useable, and less convenient to users. Including the signfiicant restriction about having an amazon account that has an Amazon purchase in it’s history.
But of course, the main takeaway here is that relying on these external services developed to serve the interests of their corporate hosts and not us is inherently unreliable. They can change anytime, the host can decide to deny us access at any time, and the hosts decisions of where to take things is going to be made based on their own business interests, which may not be compatible with what we’re trying to do, or the interests of our patrons. This makes efforts like HathiTrust, to provide digital text services under the control of libraries themselves, very important. As of yet, HathiTrust can’t give me nearly what Amazon or Google can, for reasons both technical and content availability. But this sort of effort is very important.