Goodreads Unveils Next-Gen Book Recommendations

Multiple algorithms, 20 billion data points and the "secret sauce": world’s largest community of readers

Santa Monica, CA, September 17, 2011 --( Goodreads today launched a new book recommendations engine designed to meet the seismic changes in publishing. Combining the most comprehensive set of data on what people like and don’t like to read with proprietary algorithms, Goodreads delivers book recommendations of a caliber previously unseen.

“No one has been able to provide recommendations like this before. This is the first time a community of readers of this size has collectively shared their literary likes and dislikes,” said Otis Chandler, CEO and founder of Goodreads. “It gives us data that no one else has. For example, we have more than 174,000 ratings of the best-selling ‘The Help’ while Amazon only has around 4,400.”

Goodreads has almost six million members who have added more than 190 million books to their shelves and mark more than 100,000 new books “to read” each day. On average, members have 140 books on their Goodreads bookshelves.

Combining multiple proprietary algorithms which analyze 20 billion data points, Goodreads better predicts which books people will want to read next.

Using ‘The Help’ as an example, if a reader liked it because they like reading historical fiction and they also liked ‘The Guernsey Literary and Potato Peel Society’ and ‘A Tree Grows in Brooklyn,’ then a great recommendation for that reader is ‘These Is My Words.’ With Amazon, the focus is on other best sellers so someone buying ‘The Help’ would get recommendations for books as diverse as ‘Water For Elephants,’ ‘The Hunger Games’ and ‘One Day.’

“With Goodreads, it’s as if you combine your favorite librarian, your best friend, and a database of two million book titles into one person and ask ‘what should I read next?’” said Chandler. “We’re the Netflix of book recommendations. As members add more reviews and ratings, we keep improving our suggestions for them.”

How Goodreads Recommendations Work

While other book recommendations options are based around knowledge of a book – whether you bought it, the author, the content and the genre – and then recommend similar titles, Goodreads also analyzes the reading history of people and how different books fit into their lives.

Goodreads also looks at the relationship between books. “You can map out the connections between books – how many of them appear on the same reader bookshelves, what the topics are, when they were read – and identify the books with the strongest connections to make better recommendations,” said Chandler.

In order to provide the best possible recommendations, Goodreads gets an understanding of people’s literary likes and dislikes by asking them to rate at least 20 books. To get even more insightful recommendations, Goodreads suggests that readers add them to custom bookshelves.

“How people shelve or categorize their books reveals how they think about books,” said Chandler.

For example, Goodreads members have put “The Help” on bookshelves called “Historical Fiction,” “Friendship,” “Racism,” “Women’s Fiction > Chick Lit,” “Cultural > African American” and “Guilty Pleasures.”

“You can see the diversity of opinions about where this one book fits into people’s overall world of books,” explained Chandler. “For some readers, they view the book in historical terms while for other readers, it speaks to them primarily about racism. And for still others, it’s simply Chick Lit. People take different things away from books. It’s that wide range of reactions which is reflected in our new recommendations engine, providing suggestions which are unique to each reader.”

Suzanne Skyvara
(415) 453 4880