"Digital Book" image via Shutterstock
“Digital Book” image via Shutterstock

A new book website aims to provide a counterweight to Amazon’s growing dominance in books, by focusing on recommendations—the linking, liking, and embedding experience that drives so much online culture these days. Bookish.com launched in February, backed by three major publishers: Hachette Book Group, Penguin, and Simon & Schuster. Given its genesis as the brainchild of industry giants, it’s a little bit like Hulu for books—an effort to regain some control in an era of content gone wild.
Amazon has become, according to a recent story in The Atlantic, “the most feared player in publishing.” The proliferation of tablets and e-readers, the growth of social media, and the shuttering of retailers such as Borders continue to create waves for the publishing industry, forcing readers and writers to adjust to a new paradigm. Amazon recently purchased Goodreads, a social network for avid book readers. It caters to the small but important audience of hardcore readers, offering a platform for reviewing, discovering, and buying books. Other sites like my-bookclub.com, LibraryThing.com, and libib.com offer similar fare, but none seem to provide as rich an experience, or has comparable presence (or value) in the marketplace. As the Atlantic story says, “the publishing industry survives on super fans—book worms who read far more than most Americans, and who tell their friends what to read as well.”
Bookish calls itself “an all-in-one website that uses patent-pending technology to provide a book-centric, contextual, and personalized experience, all with the goal of helping readers find their next book.” I recently stopped by the Bookish office, in New York’s Flatiron district, for an interview with CEO Ardy Khazaei.
Bookish CEO Ardy Khazaei
Bookish CEO Ardy Khazaei

What sets Bookish apart from other book sites on the Web?
Several things. One is author exclusives. When we launched, one of these, Elizabeth Gilbert’s piece on Philip Roth, generated a lot of press.
We have a team of editors, who are like journalists. They write pieces for the website, as reporters if you will, about books and authors, collections of books, motifs within a genre, books that link to a motif. They have editorial meetings to come up with ideas, like at a magazine. In fact, it’s not that much different. At planning sessions, we see what we can get from our publishing partners, usually exclusive pieces like the excerpt of an upcoming book or a jacket reveal. At launch these came from our three shareholder publishers, but we’re now getting material from other partners as well.
We launched with 16 partners, but we continue to sign contracts with new publishers who want to contribute and be a part of the website.
So content is what differentiates you?
Yes. Another example is author profiles. We pull in content from around the Web about a given author, including a bio, photos, articles, even video in some cases.
Then there are the “Bookish Essential” lists. There are must-read lists within a category, of which we have 18 right now. And within these are sub-categories. Take History, which includes areas such as the Great Depression, presidential memoirs, civil rights, etc. Or a bigger category, Young Adult, which has sub-genres like dystopian, teen witches, coming-of-age, bullying, vampires, etc.
So editorially, there is a lot to offer our users: articles, exclusive content, recommended reading lists. We talked about this in the press as a “path to discovery.” This is how users will find their way to new books. Sites like Amazon have tags provided by users, but in our case, we have editors selecting these. It’s a curatorial approach.
So it’s a matter of editors as opposed to algorithms?
Well, not quite. We do use algorithms in our recommendation tool. In fact, this recommendation technology is one of the things we’re known for, and we’ve applied to patent it.
Some sites use a kind of “collaborative filtering”: If you liked this book, here are other books that readers who liked it bought. We wanted something different. How do you develop a technology that has an understanding of what the book is about in order to make a recommendation? Our system basically ingests and pores through hundreds of thousands of book descriptions, book reviews, and other metadata to find the book’s “fingerprint.” We call it “Deep Book Introspection.”
So by having a computer go through all this, you can find certain motifs or themes you might not find by simply looking at other books in the genre. But the machine just does the heavy lifting. At the end of the process, our editors provide strategic guidance, a fine-tuning and testing of the outcomes.
The process is in beta right now. We have about 270,000 titles in the recommendation universe. I’m hoping to double that soon, so we include about half a million titles. Of course, data isn’t available for all books, and it won’t be as rich for brand-new titles.
What marketing have you done?
We’ve done some advertising. We’re being very deliberate with that because we want to make sure it’s cost effective. We’re also seeing how users navigate through the site and what kind of users we are getting. Most people provide their age and gender, and as we gain more knowledge about them we’ll develop ad plans accordingly.
We’re active in social networking, on Facebook, Twitter, and Pinterest. We also have a deal with USA Today, in which they syndicate our editorial content, three or four pieces a week, with links back to us. In turn, they share their best-seller lists with us. They will also feature our preview reader, which lets users read samples from books on our website. It’s a proprietary technology. We’re in talks with other outlets for potential partnerships.
Books can be ordered on Bookish. How does that work?
Fulfillment is handled by Baker & Taylor. From our perspective, it’s easier to not have to handle things such as shipping books. But from a user’s perspective, it’s all happening here. It’s a standard arrangement, with everybody taking a cut of the sale.
I see you also offer the option of buying through other retailers, such as Amazon. Why?
We want to enable readers, put them in charge. Books are sold here for a good price, and we hope people will buy them here, but we understand that some people have commitments to certain retailers, and we didn’t want to hamper that. The feedback I’ve seen shows people appreciate it.
Any plans for a mobile app?
For now, the only app is our reader, which has been customized for mobile devices. We’d need to be able to justify the purpose of any further move into apps, as opposed to other things we want to build and develop, such as a personalization tool that would offer tailored alerts.
Amazon’s purchase of Goodreads has been in the news. How are you different?
A big part of what they do is to connect users with others who have similar interests, even if they don’t know each other. That’s a community kind of thing. Our model does allow social sharing, but we’re not trying to connect people; it’s focused on curation. People do write reviews on out site, so there is some reader dialogue, but overall our users aren’t discovering books based on what other readers think. Our editorial perspective really sets us apart.