By now everyone acknowledges the power of Big Data to inform decisions and help match consumers with products. But only in recent years could the publishing industry access reliable, competitive sales figures to estimate a new title’s possibilities. Although many industries broadly employ data analytic tools, publishing’s use of data has mostly restricted itself to book sales and distribution. In publishing, big data’s promise has been only partially met. Until now.
Artificial Intelligence (AI) has wide reaching applications and implications–so much so that it can be understandably difficult to see the forest for the trees. But it is technology that has real potential to soothe some of the most common aches of publishers (and the film and television industry, too)
AI is growing up. Although there’s still an aura of science fiction to the image it, the reality is that AI applications can reveal in amazing detail aspects of content that were otherwise hidden from us or seemingly impossible to put into words. When analyzed across thousands of published or produced works (both hits and flops), the results show patterns that can guide us toward commercially successful future investments and away from, well, the other kind.
Discoverability–making individual titles findable by individual buyers—is now scalable. Analytics tools powered by natural language processing can deliver powerful, updatable keyword sets for an entire backlist with the same acuity that has until now required a human touch, and it does so in far less time. AI can pull both key elements of the text–characters, settings, locations, etc.–as well as the things that people are more likely to search for when looking for their next escape: experiences. “Fast-paced,” “emotional roller-coaster,” “mind-bending,” etc.: with natural language processing and AI, these qualifications can be objectively assigned in an instant. Coupled with the talents and instincts of a seasoned team, suddenly the data-entry associated with keywords and metadata can go from a tedious, anxiety-ridden task to a chance for collaboration and exploration.
Humans are great at making connections in themes and motifs (a girl on a train, a space adventure, etc.) , but machines see commonalities across genres, such as variations in rhythm, mood, and language, that can turn out to be critical to whether readers embrace a book or abandon it.
With AI, profiling the prime audience audience and matching them to a specific book gains surgical precision. Thanks to Big Data and a decade of online and digital reading, the industry has gathered vast amounts of detailed data about its readers. Artificial Intelligence can harnesses that data and uncover which groups of readers are likely to respond to a book under consideration. One tool for this job looks at comparable titles: “comps.”
For example, you’ve probably heard some version of “It’s the next Gone Girl” or “It’s the next Girl on the Train,” “ It’s Girl on the Train meets Gone Girl” in the past five years. But what, really, does it mean to call a book the next Gone Girl or Girl on the Train? That there’s a manipulative protagonist? That it has the same gut-wrenching plot twists? AI has the capability to simultaneously compare parts of a text to itself and to other inputs–sentiment analysis, customer profiles, behavior studies, etc.– to get at the heart of what makes one book “like” another.
More than just a comparison, AI can produce more detailed, faceted comps that show why a comp is, in fact, comparable. These insights help everyone make marketing plans and design decisions with a crystal-clear view of the target.
With the guiding hand of AI, the path from acquisitions to sales can increase your bottom line. When the analytical results of a manuscript evaluated by AI that has trained on bestselling books or on a specific audience’s preferences, publishers can see just how on- or off-target it is; get a clear sense of what it would take to improve the property’s aim; see how the perfect audience is likely to engage with it; and bolster the decision making process for a greater chance of meeting goals.
Artificial Technology reduces the expense, in time and dollars, of correcting mistakes. (For example, a property that needs a lot of revision to make it timely, targeted, or trendy; one for which there’s no clearly defined audience, or that turns out to be too much like, or not enough like, the competition.) Decision makers can make better decisions with powerful data and insight at their side, adding fuel to the creative talents (and yes, instincts) of the publishing team to deliver a winner. Humans, not algorithms, are still in the driver’s seat, but thanks to AI, the road has fewer potholes.
AI in service to creativity is a human-controlled operation that can spur, not stifle, inspiration. It can reveal that gem in the submissions inbox that might have been overlooked. With the help of sophisticated data collection and evaluation, it can support the publisher’s gut feeling that this book is going to be big. Publishers are still taking a risk, but it’s a data-informed and -supported risk.
And there’s always room for that rare entry in the race (how about Harry Potter?) that speaks so powerfully to millions of readers, and becomes, in turn, its very own data model. Great stories are still being told; AI can help find and successfully publish them.