Wattpad’s new publishing division uses an algorithm to identify stories
This article was originally printed in the Feb. 11, 2019 issue of The Round Table.
Wattpad is starting a publishing division. Yes, the infamous website that literary critics scoff at is branching into traditional publishing.
Wattpad has traditionally been an online storytelling application. Its millions of users could upload their own stories in a serialized fashion, chapter by chapter. There were options for both original fiction and fan fiction. Stories could easily be discovered and readers could directly comment on each chapter, giving their feedback.
The most popular Wattpad stories have even been getting mainstream media attention in the past few years. The Kissing Booth by Beth Reekles was uploaded in 2011 and gained a Netflix adaptation in 2018 starring Joey King and Jacob Eliordi. After, a Harry Styles fanfiction, by Anna Todd was not only picked up by a Simon & Schuster imprint (with a six-figure check) but it’s screen rights were bought by Paramount Pictures. The movie, starring Josephine Lanford and Hero Fiennes Tiffin, is set to release on April 12, 2019.
With its newfound success, the publishing division is Wattpad’s latest venture. The company’s chief executive and co-founder Allen Lau also described the “Story DNA Machine Learning” technology it will be using to “take the ‘guesswork’” out of publishing. This technology will “scan and analyze the hundreds of millions of stories on the app to find themes or elements that might determine a story’s commercial success.” This is said to counteract the bias in publishing which relies on individual editors’ tastes, said Ashleigh Gardner, the Head of Publishing at Wattpad. She pointed to Fifty Shades of Gray as an example of a book that wouldn’t have been published without online support.
Let me just say: an algorithm as your main editorial vision is a horrible idea.
Now, I have nothing against Wattpad as a company. I know their reputation can be fraught and there are loads of stereotypes applied to the stories uploaded there (YA romances with problematic love interests) and the writers who upload them (teenage girls). However, that is not the discussion we are having today.
The thing I take issue with is their Story DNA technology technique and their “data driven” approach.
First of all, I doubt it will work very well. Stories are not data points. You might be able to categorize some data on the stories such as genre, word count, number of readers on the site, etc, but there are limits to what you can plug into your algorithm. We have all been scrolling on Netflix late at night and never clicked on the “recommended” titles that Netflix is pushing on us. Why? Because we understand that Ten Things I Hate About You and Insatiable aren’t the same thing. They might both be labelled “teen-movie” “romantic” and they might both have gotten a lot of views on the platform, but I don’t think any human content curator would say they’re the equivalent of each other.
An algorithm has no eye for the quality of a piece of writing. It can see how hits it gets, but as we know, the most popular story isn’t always the best written. And also, great stories don’t always have mass market appeal.
As far as the diversity issue they bring up, I agree with the fact that there is a diversity issue in publishing. According to the 2017 Publishers Weekly Industry Salary Survey, 86% of 664 respondents were white. That is a huge issue and types of stories getting published reflects this.
However, the answer to the diversity problem is not to get rid of a human editorial eye for most of the process. It is not to replace it with an algorithm. Because we must beg the question, who is programing that algorithm? A computer does not get a pass for being objective. Also if those algorithms are using past data, they could be passing on the very mistakes of individuals and readers that they claim this technology will circumvent.
The answer, instead, is to hire people of color to do that work. It is to amplify the voices of the editors, writers, publishers of color that are already doing that work. Put money into their imprints and their books. It’s not rocket science and it’s not an algorithm!
There is no way this algorithm can “take the guesswork” out of publishing. That is simply impossible. Books are art no matter how you slice it. Readers are not predictable data points and the book publishing world is changing faster than ever.
Moreover, as Laura Zats put it in her podcast, Print Run, “Guessing is the best part” and as Eric Hanes elaborated, “[guessing] is the most artistically meaningful part.”