May I Help You? – Better Recommendations for Book Buyers

sales clerk helping customerIn May 2007, R.R. Bowker, the global leader in bibliographic information management, reported that the number of books published in the U.S. grew 3% in 2006 to 291,920  new titles and editions.   Advances in print on demand technology, as well as cost effective Web 2.0 marketing techniques and soon, the ability to have book production at the point-of-sale, are trends that will present book buyers with an overwhelming set of choices.  

As the aggregate number of titles in print continues to grow, it will be more imperatie for users to have better tools to sort through the market clutter and find books aligned with their interests.  Chris Anderson, in his book “The Long Tail,” highlighted the need for post-publication filters (peer reviews, search and recommendation engines) when the available investory of products becomes – for all practical purposes – infinite.  Amazon pioneered this strategy offering suggestions for books based on a customer’s previous purchases, as well as similar purchases made by others. 

business2 - June 2007 coverIn it’s latest issue, Business 2.0 reported on a new generation of recommendation engines that promises to go a step beyond purchase history.  These new tools will instead focus on discovering attributes in product purchases that the customer may not even be aware of and presenting recommendations based on these common features.  The new tools even examine consumer click patterns to deterine whether they are simply browing, researching or actively buying a product.  They can also compare products across sites to make more targeted recommendations.  These approaches are now under active testing by such companies as Netflix, Blockbuster,,  Comcast, and iTunes.  The three companies featured included:

  • ChoiceStream – recommends products based on numerous attributes or characteristics the customer values applied to their purhcase history
  • CleverSet – analyzes product descriptions, prices, ratings and multiple other attributes to make recommendations
  • Aggreagte Knowledge – makes suggestions based on cross site online consumption and buying patterns

biofeedbackI recall my experiences in bookstores when I would wander the aisles, looking at many different books, often in wide ranging subject area, trying to find that one most delicious read.  I have had similar experiences cruising the virtual book shelves of Amazon.  I wonder if these recommendations engines could figure out my thinking patterss and make that perfect suggestion.  Probably some aspects of why we desire a particular book will always remain ours – personal, hidden from even the cleverest pattern recognition software and most comprehesnive databases.  Bust who knows, perhaps this new class of recommendatino engines may act like our externalized unconcious, knowing what we want before we are conciously aware of it. 

This entry was posted in book marketing, publishing technology. Bookmark the permalink.