Stop doomscrolling. The escape Island of love. With a little nudge from artificial intelligence, you might switch to a format that is several millennia older and far more nutritious: the humble book.
“AI can be a game-changer for how we can market and sell books,” says Sarah Lloyd, global head of AI at Pan Macmillan, one of the UK’s “big five” publishers and home to international author brands including Ken Follett. , Julia Donaldson and Elton John. “Machine learning and automation, now coupled with the power of natural language processing, opens up a world of possibilities that will allow most publishers’ relatively small marketing resources to punch above their weight.”
This is a blow that has already been dealt quite hard. Business is booming for UK publishers despite persistent claims that the toxic cocktail of social media and streaming is about to render their ancient product obsolete. Nielsen BookScan says the UK book market generated a record $2.27bn (£1.83bn) last year, its highest revenue since accurate records began. Spare is the fastest-selling non-fiction book in history. In contrast, 2023 was a tough year for both Twitterstorms and box sets, with media executives declaring that we had reached “peak television” while Elon Musk continued his consistently impressive job of destroying X.
Backlist monetization
In this situation, the emergence of mature marketing tools powered by generative artificial intelligence may be the boost publishers need to win the minds – and pounds – from their shinier competitors, and they have a powerful hidden weapon: their backlists. Founded in 1843, Pan Macmillan has a portfolio of award-winners, former bestsellers and two-century-old underrated gems just waiting to be rediscovered by booktokers. The other four major players—Penguin Random House, Hachette, Simon & Schuster, and HarperCollins—have equally rich archives to draw on.
“More than 100 million. [book] titles were only published in English,” says Nadim Sadek, founder of Shimmer AI, a London-based artificial intelligence advertising platform that has amassed subscribers from more than 120 publishers, including members of the Big Five, since launching six months ago. and smaller independent publishers, about one per working day. “No more than 5% of them were ever advertised. Accordingly, they are largely under-monetized.”
Courtesy of Molly Flatt
Tastes of matchmaking and titles
Shimmr uses a three-step process to turn a simple PDF or ePub file into a piece of art you won’t be able to resist. First, “Analyzer” uses a variety of large language models (LLMs) to determine the structural patterns and psychological values of a work, or “BookDNA.” The “Generator” then uses a set of diffusion models and other image generators to create images and copies that are closely related to this BookDNA. Finally, the Deployer distributes content through media channels where readers with relevant psychological predispositions are active.
“Reward models ensure that the campaign continually self-optimizes, finding the most effective combinations of BookDNA, images and text across different channels and for different audiences,” explains Sadek. “Readers who are exposed to more psychologically relevant books also enjoy reading more, which positively reinforces their reading (and purchasing) habits.”
The results are already impressive. “In just six months of launch, Shimmr has doubled ROAS (return on ad spend) compared to the Google average,” reports Shimmr Chief Revenue Officer Brooke Dobson, “and we are on track to improve this metric by at least 3x with our self-optimization system.” Shimmerra. Additionally, game sales on Shimmr are growing at an average of 35%.”
Unsurprisingly, the company has attracted some high-powered investors and publishing consultants, including former Macmillan CEO Richard Charkin and former Penguin Random House managing director Macus Dole.
Balance between innovation and ethics
But current industry leaders remain wary of experimenting with AI, and for good reason. Dozens of writers, including John Grisham and George R.R. Martin, are currently suing OpenAI and Meta for teaching LLM copyrighted works without permission or compensation while imitating AI-generated content. threatens to undermine the already meager incomes of authors and illustrators (averaging around $8,700 (£7,000) a year, according to the latest count from author licensing body ALCS).
Lloyd, currently the only UK trade publisher with artificial intelligence in its title, is well aware of the tension between innovation and ethics. “Publishers have to meet two sets of parallel needs: on the one hand, to innovate and take advantage of AI to make us more efficient in selling the work of our authors, and on the other hand, to lobby and actively promote the use of their copyrighted materials to be respected , be recognized and compensated.”
Sadek is also keen to highlight how Shimmr protects the creators it serves. “Books are hermetically sealed away from training kits [our] models,” he insists. “Each execution consists of the original single pixel reported by BookDNA. No pre-existing images are used as a starting point.”
However, Lloyd is also optimistic about the need for book trade leaders to actively embrace and shape AI. “Recognizing the scale and impact of AI – and the power it is likely to have in the future – Macmillan (and our parent company Holtzbrinck Publishing Group) recognized that a centralized role and cross-functional approach was vital,” she explains . “AI’s potential to be disruptive means the publishing industry cannot risk being left behind.”
So the next time you’re looking for a new series to binge and find yourself buying a decade-old novel that suits your tastes surprisingly well, take a moment to celebrate instead. You’re helping the publishing industry beat AI at its own game.