[Web4lib] Google Books and Amazon - influence on book sales?

Jonathan Gorman jtgorman at uiuc.edu
Sun Dec 2 11:51:37 EST 2007



I think this would be a subtle and tricky thing to really measure.

There's the most immediate and practical benefit which would be making a person comfortable enough with their purchase to actually buy an item.  I suspect this does happen, but there's some other advantages to consider.

The first is just the fact it's there.  Sometimes it can be worth having features that are rarely used just for the fact it helps build a conception of Amazon and Google as knowing the most about an item.  Think of the technique of having comfy furniture in a store.  People aren't likely to use it frequently or consciously choose a store because of it. However, after visiting a store with such furniture they might have a better sense of comfort or security.

Another advantage is that some of the information could be use internally as well.  The text stats of Amazon is a good example here.  I imagine part of the recommendation algorithm could incorporate reading levels at some point.  (If it doesn't already).  If there's a statistically significant trend in books purchased and reading levels, you could not recommend any books too far above or below it.  Or, if you keep track of a person's reviews and all items over a certain reading level are negative, that can be feed into an algorithm of areas to avoid.  Since you have the information already, why not make it available for those who are willing to dig through a couple of screnens for it.

I would also wager some of the parts of inside the book are actually test grounds for future work by Amazon and their developers.  They're picturing an age of increased access to digital versions of things that they want to be the best at crunching and extracting what's needed to sell to people.  And it's something to keep their various employees interested.


The biggest advantage is of course it is a way to attempt to get over one of the biggest hurdles of the Internet.  There's no way to physically examine the item, so the more information given the better it can reassure people that they're making the right choice and cut down on returns and complaints.  Customers are reassured, staff time is cut, and savings can be passed on to the customer.

Measuring some of these other benefits would really require some well-done human testing.  You'd also have to find a good pools of controls and in a day and age when half of Internet users have visited Amazon in the past six months that could be difficult.



Jon Gorman


    

---- Original message ----
>Date: Sun, 2 Dec 2007 11:34:41 +0100
>From: "Anders Ericson" <anders.ericson at norskbibliotekforening.no>  
>Subject: [Web4lib] Google Books and Amazon - influence on book sales?  
>To: Web4lib <web4lib at webjunction.org>
>
>Both Google Books and Amazon have since 2004 offered "Search Inside"
>digitized books. Has anybody seen statistics or qualified guessings on
>the alleged positive influence on book sales? (Apart from their own
>statements)
>
>
>Anders Ericson
>_______________________________________________
>Web4lib mailing list
>Web4lib at webjunction.org
>http://lists.webjunction.org/web4lib/


More information about the Web4lib mailing list