Feeling confused and afraid by the prospect of using AI in your writing career? In this session, Holly Payne, CEO of Booxby, sets out to demystify artificial intelligence and reveal how it can benefit indie authors by amplifying unique voices. Plus, a look into Booxby itself and the tools available to grow your author platform.
This is a post from SelfPubCon (The Self-Publishing Advice Conference), an online author event, run free twice-yearly, in association with the Alliance of Independent Authors.
Find more author advice, tips and tools at our Self-publishing Author Advice Center: https://selfpublishingadvice.org, with a huge archive of nearly 2,000 blog posts, and a handy search box to find key info on the topic you need.
And, if you haven’t already, we invite you to join our organization and become a self-publishing ally. You can do that at http://allianceindependentauthors.org.
Listen to Authors AI with Holly Payne
Watch Authors AI with Holly PayneHolly Payne, CEO of @Booxby, sets out to demystify artificial intelligence and reveal how it can benefit indie authors by amplifying unique voices. Click To Tweet
Read the Transcripts: Authors AI with Holly Payne
Holly Payne: Good afternoon. Thank you for inviting me to your conference. I’m Holly Payne, the CEO and founder of Booxby, a cloud service that leverages artificial intelligence to aid book discovery.
First, I want to acknowledge the colossal challenges we’re facing in these uncertain times, and part of what I wish to discuss today is how to remove the uncertainty and discomfort around your understanding and use of AI.
As a fellow author, I applaud your courage to take the independent publishing path and I celebrate your creativity. The world needs your gifts, now more than ever, and you’ll soon learn today that AI can actually help you.
So, I’m recording this talk from the Booxby office in Sausalito, California. It’s about 96 degrees here today. This is my first recorded presentation, so I thank you in advance for your understanding any possible technical glitches that could happen.
In the next 40 minutes or so, I’m going to share with you three ideas. First, really quickly, why I as an author and using artificial intelligence. Number two, what artificial intelligence and machine learning is and does. And number three, how AI can actually support you as an independent author.
So, firstly really quickly, how did I, as a novelist, go high-tech? Well, I’m here because I love stories and I believe stories are the most powerful medium in which humans relate to each other and the world. I also believe the best stories come from the inside out. From the inside, we learn what really matters. So, let’s start there, the inside.
So, when I graduated from college, two weeks after college, which was many, many years ago, I was struck by a drunk driver on a dark mountain road in Colorado.
He left me unable to walk unaided for nearly a year, and I spent the next 15 years in a battle with forgiveness, but processed it all in my third novel, Kingdom of Simplicity. So, writing was my way of righting this tragedy. My femur was crushed, and my hip and pelvis were fractured. I spent weeks in bed, and eventually used a wheelchair, graduated to crutches, and then a cane, before I walked again with the titanium plate and screws that hold me together to this day.
Yes, part of my body is artificial, but it’s enabled me to walk unaided, compete in triathlons, race mountain bikes and walk my daughter’s puppy. I cannot do this without the technology that lives in me and augments my everyday life. So, I share this story as an analogy with all you storytellers, for how artificial intelligence can augment your independent publishing journey.
So like you, I’m also an independent author, but before I went down this path, I was published traditionally. My first two novels were published by Dutton and Plume, which were imprints of Penguin at the time. Back then I thought I won the lottery, but my experience was actually less than optimal. I grew increasingly frustrated with the marketing efforts, if any, from my publishers, and that led me to start my own press. I wanted to learn the industry from the inside out.
And I was told, as I’m sure you might’ve been told, that writers write, but today that’s not enough when we all know it. If we want to succeed, we have to learn the business too, like any other artist. And as a trained journalist, I saw this as an investigative opportunity, and I learned a ton.
From what I knew about setting up a title in this industry, I realized it would be humanly impossible to make intelligent marketing decisions now, and in the future, with so much content and metadata. There had to be a way to understand how my book fit in the family tree of books, and use that information to guide the positioning and marketing my work. I didn’t spend years of my life dedicated to the craft to let my books die on a shelf.
As one agent so crisply stated before I chose to leave traditional publishing, I wonder if any of you are chuckling because you might’ve heard the same thing, if anything, and this is heartening, books no longer have a shelf life. Thanks to digital publishing and AI, it’s now possible to keep your book in the market for years. I’ll share more about that, but a little more backstory first, so bear with me.
When I met Josh Convisor Booxby’s COO, he told me he was close to never writing again. Now, his publishing experience was with one of the top publishers in the world, and it was also less than optimal. So, like me, he felt very frustrated, and I have chosen my words very carefully because you can fill in any other words in between here. Frustrated is probably the minimal feeling that we were all experiencing.
Josh is an incredibly talented writer, and I sat across from him at Juice Ranch in Santa Barbara, and I felt my heartbreak. Nobody really talks about this dark side of the business. We put our heart, sweat and tears into these books, but too often end up crushed because our work didn’t find its audience. So, we’re told the market’s too small or that we just don’t fit in. And many of us quietly suffered this vicious cycle. I did. But it no longer has to be that way. There are tools to help you succeed now.
Let me tell you, there is an audience for your work. I firmly believe that, or I would not have given the last six years of my life to being a part of the solution. Maybe you have felt the same heartbreak and frustration, but you’re not alone.
So, if Josh and me, and people like you, who love the craft and are truly dedicated to writing books that serve the world, I was wondering what the world would be like if we all just stopped putting books into it. That’s a scary thought, and I don’t like that picture.
I was struggling to find a solution or a tool that would help me, Josh, and you, reach our audiences. You see, as much as I hoped to create a better experience for myself with my next two novels as a self-published author, I faced the same challenges of every other publisher, big, small, or midsize; digital publishing was making it harder and harder for me to find my readers. This third wave of disruption flooded the market with a ton of content and the landscape changed forever, as you know.
So, I wasn’t out of the woods yet. In fact, I had just entered it. So, I want to share some data for historical context. And if you bear with me, I’m donning my prior professor hat here to give you an overview so that you do have an idea of why we’re at where we’re at in the industry.
So, the number of books published annually in the US has exploded by 400% in the last decade, according to (inaudible) reports, and with more than a million and a half books published each year, publishers are finding it increasingly difficult to find readers for each book, as we know, making book discovery the publishing industry’s biggest problem. This creates a significant marketing challenge in publishing.
The same signal to noise issue is happening with podcast and video content as well. Last year alone, 18 million podcast episodes, and 116,000 streaming video shows entered the market. Content creators across all three channels are struggling with this massive marketing challenge. And as consumers we’re faced with it too. How do any of us effectively market our content so that consumers know what to read, listen to, or watch next?
This is why Booxby exists. I wanted to know if there was a more efficient and intelligent way to answer the three most critical questions that guide any marketing campaign; who would love my content? Why would they love it? And how big is the market? With so much existing content, answering these questions has become a big data problem, but hopefully, and as we are about to discuss, one that AI is uniquely suited to solve.
So, the National Science Foundation recognized this need and our innovative solution to apply natural language processing and machine learning, which I’ll explain in a bit here, to the story industry, and they awarded us a grant to build our first models with books. Specifically, novels, which represent the largest data set of long form narrative.
I never thought in a million years, I’d be using artificial intelligence to solve this discovery problem, but the age of machine learning is upon us. It’s here and it’s changing the way we will acquire, market, and discover books forever. Some call this the fourth wave of disruption. I like to see it as the fourth wave of opportunity, because technology, as I shared with you, has mobilized me from the inside out.
So, let’s take a moment here to demystify AI – Artificial Intelligence. Machine learning is actually a branch of artificial intelligence. The best way to conceptualize machine learning is by teaching a computer to recognize patterns and relationships across enormous datasets. So, why is this important?
Well, it’s a little bit like understanding genetic ancestry, current and past, to understand where you come from and where you belong. It’s about finding your tribe. Before machine learning, it was humanly impossible to track all this information, but our industry of all industries needs this augmented intelligence, and not to teach computers how to write books. Absolutely not. I’m not interested in that.
Booxby’s mission is to amplify the unique voice of you and every content creator so that you can reach your audience, and bear with me as I share a little bit about the technology behind machine learning. So, at Booxby, we’re using the science of Natural Language Processing, otherwise known as NLP, to understand multiple dimensions of books in order to help them find their audience.
So, NLP is based in linguistics, and linguists are the brave folks who can mark up a sentence and identify all parts of speech and phonology or rhythms, to name just a few things that they do. You might have remembered the agony of doing this grammar exercise in English class.
But over time, with computers, that profession became a computational linguist; one who runs a coding language to teach a computer to read and measure words without marking up the paper. That skill of identifying all aspects of written work were then applied to enormous data sets, so that the computer could do the work faster and ascertain patterns.
The process involved parsing the content and then turning that content into mathematical representations. In other words, it takes words and changes them into numbers. And those patterns over time can then be used to inform other matters. So, it’s a little like Pandora. You remember how this issue plagued musicians? Well, Tim Westergren, who was a musician himself for 20 years before he went down his tech path, he started his music genome to understand the relationship between these artists, and that’s exactly how Booxby began. I applied the same approach to books. Turn the words into math and help amplify your unique voice, because I knew you had an audience hungry for your work.
In our case, we wanted the Booxby brain to match books with similar writing style, because readers told us it was a critical factor in them choosing their next book. During our year working with the National Science Foundation, we interviewed and surveyed hundreds of avid readers. And here’s the craziest statistic, maybe not surprising to you, especially from a consumer perspective, but 84% told us they wanted books that suited their tastes and their moods. Taste and moods. So, when drill into that, tastes actually translated to, they wanted books that suited their writing style. They wanted books, not only suited their writing style, but what were they in the mood for at that given moment in time.
So, those were two key signals that we needed to software. So, with this understanding, we designed every model and output focused on these two aspects, so that the data we offer helps you tell your marketing efforts to these readers top criteria, because why put the effort into a marketing campaign if you’re actually not creating signals that will attract your readership?
So, I share all of this with you to dispel the artificial in AI, I truly wish they had called this Augmented Intelligence from the outset, because behind every model ever built is a human who is setting parameters in a code and using a data set that is also critical for results.
If you’re interested in learning more about this technical side of artificial intelligence, I encourage you to read up on data science, which is the key to building and deploying AI.
So, I’m going to take a moments’ pause to talk about some current solutions and ways that AI is being used currently in our industry, and I also want to take a moment to really be clear that we’re using AI to help amplify the unique voice that you bring, your unique combination of 26 letters that only you alone own. We’re not here to use AI to map that so that other people can then clone your DNA. That’s not it at all. We want to make sure that your unique work finds its path, and I know, and I trust that it has that path, and it has a hungry audience.
So, I want to read to you a quote from Thomas Davenport. He’s the author of, Only Humans Need Apply: Winners and Losers in the Age of Smart Machines, because it really captures what we’re trying to get to here today.
The publishing world, he writes, is full of lore about what sells and what gets read. But precious little of the lore is informed by data and analysis. So, here we go. So, publishers have been and using reader analytics for many years, as we know. Companies like JellyBooks and Inkitt engage readers, and then they analyze the book experience by embedding tracking software into digital advanced reader copies, ARCs. The tracking software is activated by roughly 300 focus group readers who sign up to get free books in exchange for providing information about the reading experience. But here’s the thing, while the insights gained are really helpful, the challenges with this process are timescale and sampling bias, as you can imagine.
And then on the consumer side, metadata and collaborative filtering have been the key techniques for solving book discovery. Unfortunately, neither offer personalized recommendations. For example, customers who bought Y also bought X, that analysis doesn’t understand why anyone chose X or Y.
Of the avid readers we interviewed, none of them were satisfied with Amazon’s book recommendations and most distrusted or ignored them. And now we understand how those recommendations are actually made, a lot of those banner spaces, that’s paid for by publishers, so those books don’t actually have a true mapping. There’s no mathematical correspondence between them. It’s been paid in, essentially, advertising space.
And while helpful in many pursuits, this technique doesn’t generate recommendations specific to a reader’s (inaudible) taste and mood. So, just because I purchased a guilty pleasure book doesn’t mean I don’t want something more challenging on my next read, but no one knows that, no one’s solving for that right now. These algorithms don’t generate recommendations relevant to what I want to read now or what you want to read now. So, without understanding why consumers read a certain book, it’s really difficult to build an effective system that makes deeply personalized recommendations. While purchase history may help understand consumer behavior, it doesn’t understand the product itself, in this case, the actual text.
So, machine learning makes it not only possible, but probable that the next frontier for acquiring, positioning, and marketing books will be informed by the analysis of the text itself, and that’s what Booxby does. We uncover the literary DNA of your work to see where it fits in the larger publishing landscape.
We’re not telling you how to write or jamming your book into a cookie cutter-based formula. In fact, we don’t even focus on looking at and understanding best-sellers; they’re part of a dataset, but absolutely not representative of all of it. We wanted to make sure we had a super balanced data set to really understand the breadth and depth of what’s out there so that we can focus on finding the markets for your distinctive style, because we believe there is a market for everyone’s unique voice.
So, I would like to share with you a quick introduction to Booxby Analytics, our platform, and show you how AI can go to work for you. We at Booxby believe every story has an audience, and we’ve built this AI-powered platform that analyzes books help you and other publishers reach the widest audience.
From the words themselves, we find patterns and relationships to hone marketing insights and maximize sales. We developed Booxby Analytics to help creators reach their audience, and it’s a tool suited for both traditional publishing and for self-publishing. So, if you bear with me a moment here, I’m going to share my screen and I’m going to get you into Booxby Analytics.
And I really hope that you’re going to be able to see this, because I can’t tell, and I can’t hear anything right now. Okay. So, this is the portal to Booxby Analytics, and we’re currently in beta. We launched our V1 model in February, right before COVID, and that was an interesting time, obviously, for everyone, and then three weeks later, the world went sideways, at least here in this country, and we’ve spent the last six months taking all that feedback from the V1 customers and created V2. So, this is based on a much, much larger data set, and we took all of that early feedback to create this version, which is our Beta 2.0, which we will be soon launching later this month. I welcome you to sign up, we’re offering the Booxby beta for free so that you can share feedback and help us make it better, so that everyone has a great experience with this tool.
So, you sign up, you get your account and, this is just an overview that captures one of the books. So, I’ve uploaded four of my novels. So, here’s my list, and what we’re focusing on, as you can see, are these three outputs. We’re focused on comparable titles, mood tags, and market opportunity, and just really quickly, at highlight level, the comps are what are capturing and answering the question of, who’s going to love my book, because it shows you this market opportunity based on it gets you beyond just the singular genre. Mood tags, answer why, this is kind of what’s in it for the reader. What’s the experience of all these words for a general audience, and then new market opportunities shows you, you know, how many copies can I realistically sell? And those are those three questions that we need to answer if we’re going to effectively put our books into the market and compete against all this other content that’s out there, not only with books, but now with podcasts and video as well.
So, let’s see, which one should I do? So, I think I’m going to do Damascena. So, Damascena is my last novel, and it was published in 2014. And as I said, you can do the math. I’ve spent the last six years halting putting any more books in the system, because I wanted to solve this issue of discovery.
So, Damascena was actually also one of the catalysts for me starting Booxby. These first two books were published by Penguin, and then Kingdom of Simplicity and Damascena were published by my small imprint, Skywriter Books. When Damascena came out, it’s a work of historical fiction, but I wanted it to read really fast. I thought I was running a ghost story at first until I realized, oh, this is not a ghost story, and I won’t give anything away, but I realized very quickly it wasn’t a ghost. So, I can’t tell you anything else, but there’s a lot of mystical elements to it. It’s very spiritual. It takes place in the 13th century during the time of Rumi, and it’s about a little orphan girl that is gifted with the ability to turn roses into rose oil and perform miracles, and she wants nothing to do with this. She just wants to find her mom.
So, that said, this book hits all these different experiences, but everyone in New York was telling me, we don’t know how to market this. It’s literary fiction, it’s historical, it’s suspenseful, it’s spiritual, it’s mystical, it’s, you know, ” all over the place”. So, we can’t really help you. And that’s when I stopped. And I said, I’m not doing this again. Why can’t I put content into the world that is coming from me, which is free and clear and pure of all these other constraints that everyone’s telling us.
I want all of you to keep on your path of creating original content, because Booxby’s going to help you take care of the rest. We want to make sure that, that original content actually finds its readers and that your gift gets received. So, having said that, I’m going to check out the comps here for Damascena.
So, remember Booxby’s comps are based on writing style similarity. So, when you get your first set of comps here, we’re showing you the stylistic comps. So, what you might find is, oh my gosh, these are in different genres. There’s one, there’s fantasy, epic. That’s kind of interesting because there are elements of fantasy and it is a saga, it’s epic in terms of the scope of time that’s taking place in Damascena. So, one of the ways that you can use the tool is, once you get your top stylistic, which are basically your ancestry in terms of whose voice is most similar to mine, which means that there’s a probability that the readers of all these other books are also going to enjoy reading my own book. So, you can add a genre filter here, and these are all the books also that are still comping on, based on that unique voice that’s in my book, but I can click on any one of these genres to go really, really granular, especially if I’m using this to find an agent. Let’s just click on, let’s go historical. So, go to historical. So, these are all the historical novels within the stylistic comparison that are mapping onto my book.
So, now I’ve got all of these “ancestors”, and most of these authors are still alive, and this gives me an opportunity to reach out to them prior to my campaign. I can, depending on when you use the tool, I have an opportunity to reach out, participate in their community, get to know them, and maybe there’s an opportunity to ask them for a blurb and let them know, Hey, we’re mapped, we’re connected in terms of our voices, and I think your readers might enjoy what I’m doing too. And obviously, we all know how social works, it’s about give and you get. So, how you show up and the proper etiquette, there’s some interesting potential that can come out of having early access to these kinds of comps that open up your readership and you wouldn’t even have known them, because it would have been humanly impossible to make these connections without an enormous dataset and an AI at work for you.
So, one of the ways that you can do this, obviously, is the comps, they are the most critical data point that you are using to position your book, and if you don’t get this positioning, correct at the beginning, that marketing campaign can really go south fast. And this is what everyone is talking about. Comps. The challenges, most of us humans can only keep track of so many books. So, what we do is we have an incredible bias, and we think, oh, what are the top books that sound good because they were best-sellers, and I’m going to say that my book is comparable to that. Well, there’s not exactly a truth to it, because if you boil it down and we look at what the consumer, the readers want, why wouldn’t we make sure that, that campaign is there to actually end up making sure that, that connection between the book and it’s probable reader is there.
So, I’m going to click on next, the mood tags. So, now know who will like the book. I can look at the comps and I can look at different genres there. I can see that my book is mapping onto obviously many, many different genres, but now let’s move onto the next critical question. Why will they like it? What experience will my book offer to my audience?
This is the second most critical data point readers told us that they need. So, you can think of mood tags as a mini review for your book. They give me a sense of what my readers will think before my book even hits the shelf. So, now I can start using these to brand the experience of my book. It’s not just so much about the content and the story, it’s about that combination of the plot. My combination of 26 letters is creating this overall experience. In those words now, I can start branding in my marketing copy, on my website, in social, in any way that I’m writing about the book. In ad campaigns. I’m going to start using these words to associate them with my book, because, the readers, that’s one of the values they are going to get out of investing hours into this reading experience.
So, you can think of them as mini reviews for your book. They are also helpful in generating keywords for SEO, and also the publishers that we worked with and piloted, they told us that these mood tags were helpful in really informing cover design, to make sure that the elements of the cover actually invoke this subliminal thing, which is the experience that you are intending for your audience. So, we’re really excited to create this aspect of the tool, because it starts to become very searchable. You can actually use this platform to search on all the other books and start to see, oh, there are those comps, and you can start to create your own exploratory path here with opportunities to engage your audience.
So, the next thing we’re going to move into is the market opportunity. So, now we know who’s going to like the book, why they’re going to like it, and now we’re going to find out what is the market size. So, to just let you know right now, our market opportunity models are based on sales data, and it is still learning. It’s part of the aspect of the tool that we’re still training, but what we wanted to do is really offer a realistic snapshot to authors, because I think so many of us come into this thinking, I’m going to sell a million copies. If we go into it understanding that the average book sells 250 copies in its lifetime, now we kind of dial back and manage our expectations. But the great thing is, we also are given a number that allows us to set realistic sales goals. And even more importantly, it helps us to set a realistic print run, because 80% of books are still done in print, 80% of consumers choose print. I know it’s an unbelievable fact, but it’s true. And 20% of books are eBooks.
We saw that trend starting to fall over the last couple of years, that people want to get away from their screens, and books are the last place; paper is the last place you can be where no one’s trying to ping you or do something to take you out of that magical and very sacred space of immersive time that we are competing against at any given moment.
So, this allows me to make sure that I’m not over printing. My first book that I self-published, which was Kingdom of Simplicity, I printed books. I had to pay for fulfillment, and warehousing, and shipping, and it was a real challenge for me, and the second book, with Damascena, I just did that through Lightning Source, with print on demand, and that helped me, but also, I don’t make as much money on each book, as you guys know, if I do a large print run. But if I get that print run wrong, it’s a terrible thing. And many publishers are plagued by this, where unfortunately, that decision for print run is placed upon the editor, and that is a very tough number to come up with, and we wanted to find a way to inform that, put some intelligence behind that number so that we can stop over printing.
There’s an article called Publishing’s Dirty Little Secret, that publishing often destroys or pulps up to 40% of its own inventory. I don’t know of any other industry that does that, other than the food industry, only because it’s perishable goods, but books should not be perishable. Let’s hope.
So, now that I have an objective analysis of who will like the book and why, and how big my market is, I can start marketing my book and know that my efforts will be focused on the right readers and give me the greatest chance of success.
I’m going to be applying the Booxby data to my first two books, I got the rights back and now I can do what I want with that content. And I have republished my second novel, The Sound of Blue, through Skywriter Books, so I’m excited to use this and setup some campaigns, and see how the data performs, especially with the V2 models. We’re really excited about the data set.
I want to share with you one other thing that we’ve created. So, during COVID, just really quickly, this is also the beauty of artificial intelligence, we wanted to find a way to help. We knew that bookstores were closing, and libraries were closing, and author tours were being canceled, left, and right, and we wanted to do something to help. What we did is, we extended our platform. Booxby Search is a free book recommendation tool, and it’s on the same platform. We’re a startup, so we were trying to stretch our resources as best we could, but we dove into this because our very same algorithms that are used to give you these tools, can be used on a consumer basis so that we can help readers find your books.
So, this is a free book recommendation tool, and it’s specifically based on stylistic similarity so that it’s coming in with that first criteria that the readers told us they needed, which was they wanted books that suited their tastes.
So, the whole goal of Booxby Search is that we hope we can serve up books to people, where they would have never even seen these connections if it wasn’t for the AI at work. So, I’m going to do a little example here. So, these are some of our favorite books. All the Light We Cannot See is one of my all-time favorite books. I’m going to show you what you do, you click it on, and these are top stylistic recommendations based on Anthony Doerr’s novel, which by the way took him 15 years to write, so keep at it. Now, this gives me a whole other new reading list, because I’m so curious to read these books, and you can also filter by genre, and you can find other books through this, but this is all because of the AI’s at work. It has read all these books, it understands and maps their relationships, and when your book enters the system, you too will be able to understand where you sit in the midst of all these other books. It gives you that place, it’s kind of like your star, where’s your constellation, which constellation do you fit in? And we’re really excited to release that. So again, this isn’t released yet, it’s in beta and we’re going to make the announcement later this month. We’ve got a lot of things going on in October, and it’s exciting to be here with you at this conference.
So I’m going to stop my share and just in closing, I just want to let you know that the world really needs your books, and we’re at such a strange time in the world, and if you can take a moment to just breathe in to the changes and the challenges and know that there are opportunities in this moment of, let’s call it intense friction.
But the world does need your books to find their intended audience. You didn’t spend all this time and pour your heart and soul into this to have it just sit there, and we want to help you find it, not by chance or circumstance, which is kind of the way that marketing has been done in the past. We want to put the power of integral AI, aiding your discovery.
So, in this virtual room, we write stories, edit them, publish them, and we nourish them. My team and I (inaudible) use technology to serve the industry in these really important endeavors, and I’m honored to share that path with each of you and see you thrive.
So, please drop me a line at [email protected], and hop over to the website booxby.com, you can sign up to test our beta. I will literally be sending out a personal email, because they all come to me, because I like to know from signing up and how they found us and how we can support you on the journey.
So, keep up the good fight, allow the world to receive your gifts, and thank you so much for being here. I hope this was helpful and informative, and I hope that you know that the world will not be totally taken over by robots and AI, that AI is actually built by humans, it’s guided by humans, and if we have integral people behind those data sets and those models, we can do a lot of good with it.
So, if you have any questions, feel free to ask me. I wish we had a chance to do a Q&A, but I’m really grateful to have the opportunity to meet you and share this. I hope it’s helpful. Take care.