The problem with Amazon KDP is that it moved the slush pile online. Harsh but true, and given the poor selection mechanism (matching customers with products they really want) you often end up with something that sounded like a great idea, but where the execution is mediocre to poor, and the customer has bought something they didn’t enjoy much.

At best it sits on the reader’s kindle, not being read. At worst, they get sick of paying for things they didn’t enjoy much, and turn to some other form of entertainment. The traditional publishing industry have mostly been fairly useless for 30 years, but originally their primary function was to filter out the slush. If it made it onto a book-store shelf, it had been curated to a reasonable degree. Then they let ideology and their preferences (not readers preferences) dictate. Then they let agents do that for them, and that exaggerated and multiplied the effect — because you had a class of NYC ‘elites’ loftily assuming their tastes and agenda were something they could dictate to the hoi polloi – most of whom they despised and knew nothing about.

It was behind the rise in independents success. But the new Indy class – often catering to the market trad had neglected by their ignorance – has a slush problem of zero curation, and worse they have an incoming AI slop problem: which could be a tidal wave that further drowns them on.

Of course: slop and slush are theoretically rapidly dealt with. Amazon has the heuristic data – every time that a reader consumes the product. How long they take to read a book/page on average, how fast, how often etc. They have a ton of other personal data for each customer. Accurate matching is very possible. Happy customers purchase more.

I am not sure they have the appetite for the results though. And when AI spreads into movies (soon) and voice acting (audio) and every book turns into an audiobook and movie… I don’t know what happens.

I suspect some new forms of curation will get born. I just hope it makes more sense than Trad pub. But I have no doubt that somehow the cream will manage to survive and even flourish. Interesting times.

10 responses to “Slush and Slop”

  1. Amazon has the heuristic data – every time that a reader consumes the product.

    Amazon also has a trust problem: there’s a more than reasonable suspicion that they have trad-pub inclinations for censorship. We know that they’ve removed books from their Kindle library without notice (even the ones you thought you “bought”) because of complaints from the “wokerati”.

    1. I’m only aware of Amazon removing a book that they didn’t have the rights to, and that only the once (1984 I think, about 10-15 years ago?). Have there been other instances?

      1. The late Colin Flaherty is the author I am personally familiar with; he published 4 books, I bought all of them, and one day the first 3 were no longer available at Amazon, and were removed from my library with no notice. He was writing about our two-tier racial justice system, with lots of examples, during the 90s.

  2. “They have a ton of other personal data for each customer. Accurate matching is very possible.”

    Yes, and they do that accurate matching with great gusto everywhere else these days. Just say “jumbo shrimp” in front of a television, a phone or an Alexa and you will get a Red Lobster advertisement. (Yes friends, as I’ve been saying for a ludicrously long time now, always-on appliances are always listening. This is no longer a conjecture, it is a demonstrated fact.)

    But they DON’T do it with books.

    So obviously they don’t make their money selling books. If they did, it would be easy to find new classics to love.

    I am going to guess they make their money some other way, possibly from the billions of dollars flying around everywhere that come from foreign concerns, NGOs, China, Russia, etc. When you look at the media landscape there’s a narrative being sold, the sellers are getting paid somehow.

    But it is still better than the dead-tree publishing industry who also don’t seem to make their money selling books. Because if they did, the contents of book stores would look very different.

    The only people who make all their money selling books are authors and book retailers. Successful authors, the ones with track record, an audience and a back catalog, except for the top-selling 0.5% are making less than a McDonald’s salary. And book retailers are going out of business all over the place. Barnes and Noble is arguably a toy store with some books in it, Chapters/Indigo is a knickknack/tchotchke store with some books in it. Dead-tree is failing. Possibly deliberately.

    Back to David’s point; they could do better, but they won’t. So therefore, shenanigans. If you ask me, anyway. The thought that people smart enough to create Amazon and make it #1 in the world would be so stupid to leave money on the table like that, it strains my credulity.

    Solution, obviously, is an AI sort.

    A third-party who DOES want to sell books will have to take all that data and match it up. This is technically feasible, and it might even be quite profitable. Even given the (apparently, but I’m suspicious) shrinking number of readers, a service that reliably finds and delivers your next ‘can’t-put-it-down’ read is something people would likely pay for. (We pay for Netflix, right? And Netflix is sh1te.)

    I am not an AI tech bro, I don’t know what it would take to sort the slushpile that is Amazon. But I’m going to guess it would take less than listening to every Alexa in the world, waiting for somebody to say ‘jumbo shrimp’. They are making money out of that, right? They do it as much as they can, always starving for more data.

    So, nerds, talk to your fellow nerds and see if there’s some clever way to buy/scrape/otherwise-acquire reader data and make a book recommending service. One that doesn’t exist to push propaganda.

  3. Some of the cream will survive. There’s no curation method that will get it all.

    For instance, The Island of the Mighty sank like a stone when first published in 1936. If Lin Carter had not found it when looking for works like Tolkien’s, and the copyright search had not mistakenly told him that it was public domain, it would have stayed there. And the three other books she wrote in that series would have remained in her trunk.

    And even with that dumb luck, it took until 1970 for her to hit big time.

  4. Word of mouth, recommendations from friends and family, is how I still find most of the stuff I read. Whenever I see “recommendations” for me online I go in and read some of the reviews, especially the low starred reviews, to see if it’s something I may in fact like. It’s also why I still write up my reviews on GoodReads, so that others may get an idea of what I liked, and didn’t like, about the book.

    Connections, connections, connections!

    1. Review and rate, everyone! Spread the word about books you like! Be the change you want to see in the world!

  5. I think genre (or micro-genre) makes a difference.

    One reason that I chose to write a LitRPG book is that they pour out in an unending torrent with mostly reasonable quality (defined as “someone at least tried to copy edit”).

    I believe Romance is much the same, but while I don’t mind occasionally reading it, there is no way I could write it.

    There is no way anyone is buying that many books. KDP/Kindle Unlimited is where that action is.

    I’ve only recently downloaded the first that I suspect is AI generated. I’ve downloaded some dreadful ones; mostly “author’s first book” efforts. I’m really hoping I don’t join that club. But with KU, who cares? Just return it and move on.

    I recently re-started The Case of the Misplaced Detective to see how Holmes reacted to modern life (as an analogy to how someone would react to a System suddenly showing up). I was shocked at the low review count (and left one). That’s a great series. Published in April 2018. 13 star ratings (4.7 average) and 13 reviews (that’s a good ratio).

    In contrast, Adam vs The Apocalypse was published in June 2026. It has 280 star ratings (4.6 average) and 23 reviews. It’s a fine series, but nothing particular special.

    These are not being read by the same people.

  6. As a reader and not an author I have found the Amazon recommendations to be reasonable based on what I’ve read in the past that were obtained through Amazon. I’m not looking for great literature but stories that entertain. I probably own the majority of books that fit my general genre preferences written by the authors who write on this blog. Those that are on KU I usually borrow first and at the next 2x/3x points days purchase them. I hope this gets the author more revenue than only buying.

    KU is invaluable to me as I read between 200-300 books a year. It is a kind of slush pile but the recommendations after I borrow or buy are a decent filter. The Amazon recommendations for Also Bought, Continue Series, and Other Books by an Author work well for me while the You Might Like one does not. My other source of recommendations is this and other blogs/Substacks. I don’t use Facebook, Instagram, X, or other widely viewed social media.

    The Continue Series is good for letting me know about new or upcoming releases and Also Bought usually has interesting options which require further research. Since I’m not a particularly eclectic reader they work. Books from these lists get added to my Wishlist to keep them from getting lost in the blizzard of other titles. The Sunday blog posts here give me new options when the blog authors go off into new genre which I wouldn’t get from Amazon.

    All in all, having Amazon as a slush pile that I curate for myself works well for me.

  7. I find that getting a sample of the beginning of the book off Amazon works well.  I can usually tell in that space whether a book is for me.  I can, of course, realize halfway through how dreadfully wrong I was, but I can stop reading then, too.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Trending