Holly Payne beat the odds. After a drunk driver struck her, she spent a year regaining the ability to walk. During this time of immobility, she read a lot of books and gained the insight required to view a book from the inside-out.
Holly beat the odds again, with the successful publication of her first novel. Average sales for first-time authors are less than 500 books. When bestselling books are removed from this average, most first-time authors’ sales amount to a few dozen books purchased by friends and family.
Holly plans to beat the odds one more time with Booxby, a SaaS cloud platform that she Co-Founded with Mark Bregman. Booxby uses artificial intelligence (AI) to help publishers optimize the acquisition and marketing of their books. Holly is the company’s CEO and Mark is on the Board of Directors.
Another Alchemist Shares Her Vision With The Universe
Much has been written about computer-generated prose, with some pundits claiming that novels written by algorithms will eventually prove more popular than human-generated content. Maybe, but I’m not sure that’s a world I want to live in.
Fortunately, Booxby’s mission isn’t to teach computers to write better books. On the contrary, Holly’s goal is to facilitate book discovery and allow more authors to flourish by directing the right content to the right readers.
John Greathouse: I know it’s a pedestrian way to start an interview, but I love origin stories. What led you to found Booxby?
Holly Payne: This story begins with an accident. Two weeks after graduating from college, I was struck by a drunk driver as a pedestrian and left unable to walk for nearly a year. I spent the first seven weeks in bed. That event forced me to read a lot. A friend gave me a copy of Paulo Coelho’s The Alchemist, which fundamentally altered my life, opening me to mind-bending ideas—and experiments in trusting the “unknown.” I had never engaged with a book this closely.
Until then, I had never known what it was to understand a book from the inside out. The concepts in The Alchemist invited me to see the world differently. Until then, I had planned on becoming a foreign correspondent, based in Hungary. While the accident altered those plans, the book itself completely rewired me. I was ready to follow my dreams. I moved to California to become a writer and even met Paulo Coelho at a book signing, in Menlo Park. His words gave me an internal compass that I carried with me to this day. He told me, “Holly, dare to risk.”
And I did, moving to San Francisco to write my first novel. That book, The Virgin’s Knot, became a Barnes & Noble Discover Great New Writers book. I was ecstatic—off and running, doing what I loved, until low sales for my second book almost killed my career.
For my third book, I decided to start a publishing company, to take control of my professional life, and then help other authors on their path. But that again came to a screeching halt when I realized I faced the same problem of every publisher. How could I guarantee that the authors’ work would reach its intended audience? How could I trust my instincts in marketing decisions? I could not rely on luck.
My fourth novel drove this point home. No one could see where Damascena fit in the market because it was a crossover title: historical fiction, mystical, suspense.
The rejections carried the same message: “we like the writing, but we don’t know how to sell this.” It wasn’t about the writing. It wasn’t about the story. I had written a book that didn’t fit into publishing’s structure —it hit multiple genres but didn’t fall into any single one. The old style of publishing had failed me.
I was devastated and frustrated. I decided not to put another book into the system until I could find a way to aid book discovery.
Greathouse: Wow. I’m a bit freaked out, because I have been using The Alchemist in one of my entrepreneurial courses at UC Santa Barbara for several years. I continue to use it, because each year the majority of the students express a similar sentiment. The book really speaks to many people, especially young people, on a deep level.
Sorry to interrupt. So you started Booxby as a means of aiding in book discovery.
Payne: Yes, that’s correct. I started Booxby to connect each book to the right readers. I sensed there had to be a way to understand narrative content at its deepest level through NLP (natural language processing) so I wrote a proposal to the National Science Foundation. My hypothesis: narrative books were actually an experience. If we could quantify that experience mathematically, we could fundamentally change the way books were acquired, marketed and discovered.
With the NSF grant of a quarter million dollars, we got to work. We went through many iterations and were finally able to validate our approach. Though our testing showed empirical evidence that this was all working, I felt it on an emotional level when Booxby’s algorithms made the connection between my debut novel, The Virgin’s Knot and The English Patient, which I had read twenty plus times, highlighting and dog-earring it to near destruction.
So that first accident was not accidental, leading me from journalism to writing novels, from writing to publishing, and from publishing to Booxby.
Greathouse: Now that’s an origin story! Your self-awareness and optimism is commendable. I love it when entrepreneurs are able to leverage grant funding to get their ventures off the ground. Well done.
Prior to Booxby, neither you, nor Josh Conviser, your COO and fellow accomplished writer, have any technical experience. Have you gotten push-back from potential investors, asking, ‘Why should a novelist lead such a highly technical endeavor?’
Payne: Not really, at least no one has been bold enough to say this was a concern. Learning how to build novels and screenplays gave me a great foundation for Booxby. All writing is about revision. All successful startups revise their plans until they have clarity and know what to build and confirm it serves a market.
The truth is that every effective fiction writer is a story engineer and skilled revisionist. In order to build machine learning models using NLP for the work we wanted to accomplish, it’s critical to understand narrative design. Josh and I know story building from the inside out—with a deep understanding of the text itself and an insider’s perspective on the business of publishing.
This domain expertise allows us to work with our immensely gifted computational linguists and data scientists and solve for issues that people without our background might miss.
John: At first glance, I thought Booxby was trying to rig the system by creating content that is optimized to be a “hit.” I thought, “This could lead to entertaining, but potentially vacuous books.”
I’ve since learned this is not at all what you are up to.
Payne: In fact, we do the opposite. Our AI is all about augmenting intelligence; we want to enhance human abilities, not to normalize or limit them.
In Booxby’s case, we use NLP and machine learning to understand and amplify unique voices around the world. Booxby is not interested in teaching computers how to write, or humans how to write cookie-cutter hits. We want our technology to allow humans to do what they do best—to understand the human condition and create original works. Booxby then understands how that creation fits with the works already out there, optimizing its market opportunity.
Booxby is committed to aiding original content, not derivative works, and we believe stories will continue to be the most powerful medium in which humans relate to each other and the world.
Greathouse: In the past, you’ve noted that Booxby can, “map correlations between books and create a family tree of literary DNA” using AI. Clearly this is a powerful capability. Where do you see the technology progressing, once you nail it in your initial book publisher market?
Payne: We began with publishing because its content is both copious and relatively easy to ingest and analyze. Our understanding of narrative content derived from that dataset can then be applied across the story industry for film, video, audio, advertising and much more. Anywhere that creators produce narrative content, Booxby can find patterns and predict outcomes.
Greathouse: Even the experts find it difficult to pick hits, whether it’s book publishers, music executives or film producers. How will Booxby help first time authors, in your words, “gain a voice?”
Payne: Helping authors gain a voice, and then amplify that unique voice, is the whole reason we started Booxby. There are two kinds of first-time authors – traditional authors represented by an agent who have a publishing contract, and independent authors, whom we consider micro-publishers.
Traditional publishing throws a lot of people at each book, editors, PR and marketing directors, etc. Throughout the process, Booxby’s data becomes the advocate for the author when there is no one else “in the room” with an intimate understanding of their text. We help everyone on a book’s path go beyond guts and guesswork to create a consistent campaign, connecting that book with its market.
Greathouse: As with most AI, the data sources are a huge differentiating factor. I understand you tested and hardened your platform using public domain content, such as Project Gutenberg. What were the limits of that content and what are you currently using as your primary data sources?
Payne: Building this unique dataset of narrative content was our biggest challenge and is now one of our biggest differentiators. In order to gain traction with publishers, we needed to be sure the outputs were helpful to them, so we prioritized building relevant, balanced datasets that weren’t too heavy with bestsellers.
It took a long time to build it, but we knew that Gutenberg would be insufficient when it came to commercializing our outputs. When we were working with the National Science Foundation, we purchased a license from Ingram Content Group, the world’s largest distributor of books. From that metadata, we built a list of titles that were relevant to the marketplace, and then sourced those titles using various methods, scanning copies, etc. We have also worked with some of the top publishers in pilots, which expanded our dataset greatly. Ingram became a strategic investor in Booxby this year, which gives us a huge opportunity to expand our corpus again and further refine our AI.
Greathouse: As noted above, job one is to nail a vertical with Editors and Publishers. However, what do you envision as future applications for Booxby’s platform?
Payne: While we’re initially targeting publishers, our service will be integrated throughout the industry, from authors, to agents, to retailers. And that’s just the beginning. The larger story industry deals with the same discovery issue that plagues publishing. Be it film, podcasts, or the newest VR experience, finding an audience for each piece of content is brutal when there are so many options. Booxby cuts through that static.
Film and streaming is our obvious next step. For content producers seeking work to adapt, Booxby offers a unique view of the publishing landscape. We’re in the process of working with several Hollywood producers to identify narrative IP that suits their desired market.
Greathouse: Pretty much every profession is dealing with how to incorporate AI into their workflows. It won’t be long before it will be deemed malpractice for physicians to not incorporate AI into their diagnoses.
Lives aren’t literally on the line in the world of publishing, though it may feel that way to a first-time author… but how fast do you think the relatively ancient world of publishing will incorporate AI into its daily decision making?
Payne: Like every other industry, the age of machine learning brings with it both disruption and opportunity for publishing. Given the sheer volume of books published these days and all the metadata associated with each one, the need for machine learning models is definitely not a fad.
Following music and film, publishing needs to evolve from making decisions through guts and guesswork, to operating with an understanding of the publishing landscape in its entirety and seeing how each new story fits into that web. That ecosystem is simply too big for humans to fully glimpse. By ingesting the stories themselves and making analyses from that, Booxby helps curb the biases that limit stories from finding their audiences. Right now, publishing is a $96 billion global industry, with the top two percent of books accounting for most of that revenue. There’s a huge opportunity to use AI to move the needle, even slightly, to increase the bottom line for everyone.
Already, big data analysis is being done with metadata and buying patterns, but that’s just the start. Using AI on the stories themselves, again, not to write those stories, but to understand them and their markets, is the next logical step. End of the day, this is an industry passionate about its product. We all want to find great, new books and connect them to the right reader—that’s the goal. As the number of books grows exponentially each year, Booxby is excited to serve and improve that effort.
via Forbes – Entrepreneurs https://ift.tt/2OoHmV0