A brave new world of algorithms

  • Post author:
  • Post category:Blog

This article in Slate about algorithms, discovery, and creativity really got my brain-juices going.

You see, one of the hardest things about being a small press – or a debut author, or an early-career author, regardless of how you’re published – is discoverability. Without a multi-million-dollar PR and marketing budget, how do you get your book in front of actual readers? How do they find you, amidst all the thousands and thousands of books published each year…let alone the bestsellers that are already out there, taking up most of the shelf real estate in bookstores and the viewing slots on Amazon?

Companies like Amazon and Netflix rely on complicated algorithms to suggest other titles for you to try out; companies like Goodreads and Fuzz rely on human curation and recommendation on a fairly large scale to do the same. Electric Literature has a really nifty magazine, Recommended Reading, that I wish had a solid sf/f/wtf counterpart (it can take forever go through blogs getting recommendations). Based on anecdotal evidence, combined with the stats gathered by the Book Industry Study Group and others, we’ve come to the conclusion that most readers find their next book through a combination of machine suggestions and human suggestions from trusted sources – friends, relatives, coworkers, and bloggers whose taste they like. Ads and marketing can factor in – you’ll certainly pay more attention to a recommendation from a friend if you’ve already heard of the book from seeing an ad somewhere – but reccs are the key.

Next up: the story assembly line!
Next up: the story assembly line!

Where the article expands on all this is the idea of machine recommendations, and using the data from sales points to determine what will sell next. It’s widgetizing creativity, basically – a great idea for publishers who want to minimize the risk when picking and producing a new book, but a lousy idea for artists and audiences.

Because if all we ever do is tweak a previously best-selling trope, where does the fantastical new thing that starts the next trope come from? Where’s the disruption? Where’s the fun?

Publishers are supposed to take risks. That’s why we play the game – we find something that we truly believe in, something fantastic and wonderful that just gets under your skin and makes you incapable of not thinking of it, and we want to share it with the world. We take on the financial risk of producing and packaging that idea, and then we go out and introduce it to the world as best we can. We hope that readers will find it, then devour it like we did and go on to tell others about it, spreading the idea, the story, the love to ever more fans.

When that happens, the author makes a bundle and we make back what we spent on development, plus enough to start financing other projects, other stories, other ideas.

Or at least, that’s the blue-sky, ideal scenario. (Well, okay, the REALLY ideal scenario ends with both author and publishing staff having their own custom-built castles on sky-islands, but baby steps, people.)

In the Widgetized Creativity system, a computer model crunches data from prior sales, and projects what combination of love story/vampire/werewolf/angsty wombat/snarky dialogue/purple prose/dystopia/alt-Victorian/space opera will optimize selling potential. And then we get sixteen variants on that, and nothing new and crazy and gripping really comes out of the process, because that’s too much risk. What if that existential zombie novel is too thinky? What if the noir caper is too funny? Oh noes!

Look. The idea of an algorithm-scripted novel is kinda cool, I’ll give you that. But the idea that all we’re going to have moving forward, in order to optimize profitability, is Widgetized Creativity – that’s terrifying. That’s practically dystopian.

And yet it gives me hope.

Because it means that there’s always going to be a place for small press: idealistic, slightly crazy, definitely willfully blinkered publishers who believe in the power of storytelling and character and craziness more than in the siren song of profits. Yes, we want profits – they mean we can pay our staff and pay our authors and replace that old computer that’s starting to smoke worryingly – but we can live without the custom sky-island as long as we’re bringing in enough to finance the next cool project.

Because in the end, it’s about those cool projects, the stories that get under your skin and won’t leave your hindbrain, no matter how weird they are. It’s about finding them, and developing them, and producing them, and sharing them with the world.

Not about making a widget.

This Post Has 9 Comments

  1. David Fuller

    YES! I think you’ve hit the nail on the head. It’s such a risk to engage in what is, essentially, literary R & D, but that’s where rules get broken enough to create something new. Or refine something that has depths yet to plumb, which may not happen when you have to scramble to widgetize the next big seller.
    I’m not sure the relationship between small press, big press and the huge booksellers will change that much — I think the big fish are going to try to devour as many of the small fish as they can for as long as they can — but as the duty of the reader seems to be changing, maybe the small press relationship to readers will, too. Readers seem to have more power to give feedback than before, not just buying power; and Goodreads doesn’t care whether you’re rating and reviewing an old book from the library or one you’ve just bought. Maybe the long tail, for readers, is getting longer, thanks to meaningful recommendations. At least, I hope it is.

  2. Candlemark & Gleam

    See, every book is a risk, though. Even something that seems to perfectly suit the current zeitgeist and what people claim to want might flop…I’ve seen it happen. So doing “literary R&D” is still totally worth it – it’s risky, but it could also really pay off if it taps into something that isn’t being served in the market.

    Besides, how will you ever find the next genre-busting wonderbook if you only print the same damn story with the serial numbers filed off?

    I do agree that the relationship to readers is changing – it’s becoming more of a two-way street, and I LIKE that. Reader feedback and interaction is a good thing, and the ability to talk directly to readers, instead of going through an intermediary, is going to be critical for small press and self-pub to thrive…

  3. David Fuller

    Yes, I think so, too — maybe readers will have more ways to make their hopes/dreams/preferences for something new more known to publishers (? just spitballing, here) — so presses that were on the lookout for the next book they wanted to take a chance on might also have an indication that it matched, in some way, what certain readers were looking for.

  4. Candlemark & Gleam

    I’ve seen some startups trying to do crowdsourced novel selection/publishing/slush reading as sort of an uber-version of this idea. You really need a massively engaged/active community to make it work, but I could see it being an interesting experiment for some larger houses…

  5. leeanna @ leeanna.me

    I know Amazon’s system is why a lot of indie authors try to get reviewers to post their reviews on Amazon, and to rate them highly, so their books show up in the “Recommended” feature.

    I think small publishers/indies/anyone writing quirky, new books have a better chance now than they did a few years ago, and hopefully even better luck a few years into the future. It’s something I’m excited about. I want to read new stuff, not the old recycled tropes (although those can be fun).

    I have access to a few Overdrive catalogs, and it’s made me really happy to see books from indie authors and small press publishers right next to books from bigger, well-known publishers. I hope more libraries take advantage of that to expand their collections. It gives readers something to check out when all the big name titles have waiting lists in the double digits. The OverDrive “Next Generation Experience” also has recommended books, and I wonder how they list titles to recommend. Maybe by genre?

    In the end, though, it’s awesome to know that there are publishers out there willing to take a chance to produce something new.

  6. Candlemark & Gleam

    There is a better chance at something taking off than before, since eBooks have democratized the publishing process, but there’s still a HUGE barrier to sales in the form of discoverability and people’s willingness to try new authors/books/ideas. I do hope that some of this gets ironed out soon, with new methods of finding books that don’t merely rely on Amazon algorithms and who-knows-whom (which can lead to sock puppet reviews and review mills – no good!)

    OverDrive is awesome, though. For ever so many reasons!

  7. leeanna @ leeanna.me

    I know a lot of libraries don’t have big budgets for ebooks, so you’d think they’d want more bang for their buck 😀

  8. Candlemark & Gleam

    A lot of libraries have a hard time dealing with the restrictions publishers place on “loaner” eBooks when they’re even available, though. It’s short-sighted on the part of the publishers.

  9. Candlemark & Gleam

    A lot of libraries have a hard time dealing with the restrictions publishers place on “loaner” eBooks when they’re even available, though. It’s short-sighted on the part of the publishers.

Comments are closed.