There’s a moment that sticks in my mind from an early meeting about my debut novel, The Trader of Saigon. The book didn’t fit neatly into an obvious genre and I asked my publisher – excited and naive – who it would be marketed to. “Women,” came the immediate response. “What, all of them?” I asked, half laughing. The answer was yes.
The rationale behind this makes sense at face value. Women buy the most books. They read the most. They talk about books with their friends, sharing and championing them in ways unlike men. Why wouldn’t we want to tap into that? Yet it doesn’t take long for the arguments to untwine.
A tried-and-tested cover design, for example, may sell more copies of a book in the short-term, but if it misrepresents the actual content of the novel, the book will ultimately end up in the wrong hands. Wrong hands mean disappointed readers, poor reviews and fewer sales in the long-term. That’s no good for publishers, authors or readers.
In the end, my novels were marketed at the ‘Book Club’ audience, complete with pinkish-purple colouring and images of young women gazing towards the horizon. They’re beautiful covers but were they right for the books? I’d written stories of hope but set against backdrops of war and vice. Male viewpoints outnumbered the women by two-to-one. Some of my most unfavourable reviews have come from readers who disliked the swearing or violence. Others had hoped the characters would fall in love. It’s not those readers’ fault they didn’t enjoy the books. If they’d had a better way to understand the type of story they were getting, they wouldn’t have wasted their money and time.
It’s this frustration that led me to create Bookaxe, a tech start-up on a mission to get the right books into the hands of the right readers. Rather than trusting cover imagery and blurb, we analyse a range of data points about the characteristics of a book’s content – from the pace of its action to its explicitness and the style in which it’s written – and use that to drive highly personalised reading recommendations for our users.
With the data we’ve now gleaned from Bookaxe, it’s obvious that the Book Club audience was not ideal for my novels. In general, we’ve found that those readers have a strong aversion to swearing and graphic descriptions of sex and violence, preferring instead character-lead stories that introduce them to issues in a more light-hearted, optimistic way.
Of course, we have some nifty tech to call on – but everyone from self-published authors to more traditional book marketers could benefit from rethinking the assumptions they make when they’re trying to find a passionate readership. So:
Have you really questioned whether your ‘target audience’ is a good fit?
Are you fully maximising A/B testing on your titles, blurbs and cover designs to see which combinations get the most positive engagement from different audiences?
Are you trying to fit your book into a trending genre when you’d be better off highlighting its differences, not its similarities, to the stuff already crowding the Waterstones tables?
And have you questioned whether you could be shooting yourself in the foot when it comes to reviews?
Take Yann Martel’s Booker Prize-winning novel, Life of Pi. Across all Bookaxe users, Life of Pi has a quality score of 48%, making it an all-round average book at first glance. Yet readers with similar preference profiles to my own give the book a quality score of 70% – much higher than average. Readers with a similar profile to Bookaxe’s co-founder, Scott, score the book at just 25%. Long before we started Bookaxe, Scott and I squabbled over the relative merits Life of Pi. It’s one of my all-time favourite novels, but one he loathes.
Much as with our discoveries about cover versus content, we’ve repeatedly found that the same books are rated highly by one demographic and panned by another. Just look at Eleanor and Park, the internationally mega-selling hit by Rainbow Rowell. It has more than 630,000 reviews on Goodreads. 250,000 of those reviews awarded the book five stars – but 45,000 people gave just one or two stars instead. 45,000 voices is no small number. How are potential readers to know which camp they’d join? What matters isn’t how many people say a book is worth reading, but who says it’s worth reading. Readers need to know who the person is that’s rating and reviewing a book before they decide whether to trust them.
So now ask yourself: how are you making sure the right influencers are influencing the right readers?
At Bookaxe, we’ve tackled this issue by using machine learning techniques to directly connect our users to like-minded readers, presenting the views of people with similar tastes higher than those whose tastes wildly differ. Every book is also listed with a ‘Likeminded Reader Score’, telling users not only how that book has been received by the Bookaxe community as a whole, but how it was received by readers with similar preference profiles. It means readers have the very clearest picture of what they’re going to get.
The critical transferable learning here, however, is that the readers who are panning books shouldn’t be reading them in the first place, and that it is possible to stop them doing so with smarter marketing. For authors and publishers, this might involve being more selective with which influencers you give Netgalley proofs to. It might involve experimenting with unexpected passion groups, who might spread trusted word of mouth within their communities. It might simply involve reminding yourself that the old adage ‘any press is good press’ doesn’t apply to book reviews, and spending a bit more time doing your research and targeting fewer but more relevant journalists.
We don’t just need to better match books to readers; we need to better match readers to readers. We created an entire start-up focused on these two clear aims, but it’s something that everyone in the publishing industry can improve, with a little lateral thinking and a commitment to question and recalibrate the most obvious marketing routes.