Latest Author Earnings Report

This one is a doozy. Not only because three reports taken at 3-month increments all agree with one another, but because the myth of self-publishing only working in certain genres is busted. Turns out that indies do indeed dominate in those genres, but they have taken significant share everywhere else as well.

I have indie friends who think we are crazy to urge Hachette to lower their ebook prices and be reasonable for the sake of their readers and their authors, and you can see why in this data. Indie authors are producing great works at unbeatable prices, and readers are rewarding them with market share. In this snapshot, indies as a cohort have overtaken all Big 5 authors combined when it comes to daily royalties earned on the Kindle Store. Unreal.

Also some info on DRM here. Short version: It’s worse than useless; it might be hurting your sales.

Check out the July 2014 Author Earnings Report here.

25 responses to “Latest Author Earnings Report”

  1. I love these reports, Hugh. It’s very heartening to hear about the growth of indies in genres outside those usually considered to be the only ones good for indie publishing, and to hear that DRM can be an inhibitor to a book’s success. It’s becoming increasingly clear that independent publishing is a viable route to take no matter the specifics of an author’s book.

    I do have to ask though, is there any possibility the dispute with Hachette could have affected these numbers? I imagine it wouldn’t be significant since there hasn’t been much disruption on the ebook side of things, but with the halting of preorders on some Hachette titles (was that even on ebooks, or just print?) and the media campaign to push buyers to other retailers, could there have been an effect on the data?

    1. I had the same thought. Maybe there was a correlation. But that wouldn’t undermine the conclusions.

      1. No, I don’t think the conclusions would be affected much, I’m simply wondering if Hugh and data guy think there’s been an effect, and if they have any idea how great of one it may be.

  2. Hugh, I hope you and your team will keep these going for a while.

    We are just seeing hints of a trend there (notice the Big-5 share seems to be shrinking). I’d like to see what starts to show after 3–5 years of data has been compiled. I think then it will be difficult to deny that there is a change going on in the publishing industry.

  3. But HarperCollins and Hachette have now started direct selling their eBooks! Hachette’s appear to be 12.99 across the board, HC has a range of prices, but both of them are locked down with DRM.

    I betcha Amazon is quaking.

    1. If they were serious the first thing they would do is sell DRM free mobi files. Then Amazon might take notice. With direct selling the way they are… it’s a dead duck.

  4. It’s really nice to see this data, and also to see that you’re expanding the number of books that are tracked. However, there is a serious problem with the increasing sample over time. Because you’re sampling a larger number of books in each subsequent period, your trend analysis is not valid. It is possible, and I would say probable, that the #50K to #120k range are mostly indie published books. If you want to do a proper trend analysis, do comparisons that are limited to the top 50k books for all three time periods.

    1. We are sampling every ranked book on every Amazon bestseller lists. The number of categories is growing. Our sales data isn’t heavily affected, because the new titles to appear aren’t selling in massive quantities.

      We’ve taken three full snapshots thus far. This data collection method is all of six months old. As the ebook market and listing of categories stabilize, and as we continue to pull more quarterly reports, the power of the data should just improve. Already, it has told us things that people speculated about before but we are finally able to see with our own eyes.

      1. Hugh – Just want to say thanks again to both you and Data Guy. This information is incredible. I’m really curious to see how things look once you have a full year’s worth of information at your fingertips. Rock on!

      2. I was not suggesting that the data were not very useful. It is great that you guys have gone to the effort to collect the data, and many people are thankful for it. That is not what I’m taking issue with.

        I also can appreciate that it is not a simple thing to analyze this data when Amazon keeps on changing the way that they categorize books. I am not suggesting that there is necessarily a problem with how the data is being collected. The real world presents data in ways that are less than ideal all of the time. A lot of the time, the trick is being able to tell what you can learn from the data and what you can’t.

        There is plenty of evidence that those additional books being counted matter. The 50k books accounted for 40% of Amazon revenue. The 120k books accounted for 50% of Amazon revenue. The third sample is a full 25% larger than the first sample with regard to revenue as a percentage of Amazon’s revenue during their windows. The increase in indie share is 4% or less of the total in all cases. Meanwhile, it can be estimated that the 50k to 120k book range accounts for ~20%(the 10% increase from 40% to 50% is 20% of 50%) of the total revenue during the third period. There is plenty of room in that 20% to account for a ~3-4% increase in share for indie authors. The data between period 1 and period 2 are a bit more compelling, but 40% is still only 93% of 43%. That leaves 7% of the total to account for a 1%-3% increase in share. This is before we start talking about error or significance.

        We all understand that there aren’t enough data points to draw conclusions beyond it being clear that indie writers aren’t insignificant in Amazon’s ebook market and that the aggregate take home pay on Amazon ebooks for indies is even more significant compared to legacy authors. Further, I know I’m looking forward to watching the data you so generously have helped to provide trickle out over the coming months. However, it is important to recognize when we are comparing apples to oranges, even when speculating about what the data may be telling us. Because there is a very significant increase in the percentage of Amazon revenue and book selection that is being analyzed from period 1 to period 3, a small change in indie share can’t be seen as even a promising glimmer. It would be nice to think that indie writers are gaining share, but the data as presented don’t suggest it at all.

        1. Yeah, we aren’t willing to call any of these movements “trends” until the number of books scraped stabilize and we have at least 1-2 years of data. We could have waited a year or two to release these findings and have them be more comprehensible, but we thought we’d let everyone in on the joy of discovery as we went along. Precisely because people such as yourself will come behind us and double-check everything.

          We’ve made all of our data available from day one with the hopes that people will play around with it. Some have. No one has found anything to suggest that indies haven’t taken a massive chunk of market share. In fact, we played around with pushing variables in favor of traditional publishing well past any sane point, and our conclusions still stand up. It’s a very robust sample size, and our findings can be bent, but they can’t be broken. Anyone who says otherwise is probably trying to sell you something. :)

    2. I’d suggest you check the repartition in the “raw” data that Hugh and Data Guy make available. From a rough quick look on it, repartition is more or less uniform for Indie… (But the tail seems “Uncategorized” heavy.)
      It would be great in the next reports if some kind of “sliding average” were made on the categories repartitions all along the samples data… See if for example indeed the Big5 books are more on top, Indie heavy on the tail…

  5. A very sincere and heartfelt thank you, Hugh. Your work is ground breaking, imperative, and inspirational.

    These reports enable those of us that are new to this space to make intelligent business decisions. Invaluable!

    Thank you.

    1. I just hit “publish.” Data Guy does all the work. But I accept your thanks and will pass them along.

      Yo, DG, HN Wake says, “Thanks!”

  6. Why so few increments? Is there heavy cost involved in your data extraction process?

    It’s not that the data or conclusions aren’t valid (haven’t looked closely enough to make those judgements), but everything would be strengthened by a greater data depth than a couple snapshots.

    1. The Author Earnings reports are cross-sectional studies of Amazon ebook sales.

      Each report involves sucking more than 100 Gigabytes of Amazon product-page HTML across the network at a sustained bandwidth in excess of 200 Megabit. The real-time data extraction uses 40 servers with 8 CPU cores each to process that incoming stream, and writes it to a database. The cloud-compute server costs to run the report exceed $100/hour.

      But the main reason we don’t do this more often is that we’d rather be writing our books.
      This is just our way of giving back to the industry we love and paying it forward :)

      1. Holy crap! I thought all the metadata (other than rank) was available via web service calls. You’re SCRAPING all those pages to pull out all the characteristics?! That’s both impressive and terrifying.

        I hadn’t looked closely enough to realize your methodology required scraping each one of those pages, so, apologies. I’ll keep my eye on everything as the data depth increases with each snapshot.

        Also, while I’ve got your attention, the rank to sale correlation was a pretty brilliant keystone. Kudos.

        1. The real credit should go to the pioneering indie authors who gathered and collated rank to sales data for their books, compared notes with each other, and then shared or published their crowdsourced numbers. Intrepid writers like David Gaughran, Theresa Ragan, and many others laid the groundwork.

          We simply stood on their shoulders.

  7. Thank you, Hugh. This really is awesome work.

  8. Hugh,

    Of the 1300, or so, authors who inputted data on, my personal self-publishing goals match 138 of them. [ 10.62% ]

    Out of 138 authors since 2010, who self-published from 1 to 10 books, hired a freelance editor, and paid for a professional cover to their book, 31 make a living doing so (at least $35,000).

    That’s a 22.46% chance in 5 years of making a living based on the above criteria [ 1.4% with 1 book only, 2% with 2 books, 4% with 3 books, 5%:4, 7%:5, 10%:6; 14%:7, 16%:8, 19%:9, 22%:10 ].

    Already, this is enlightening to me. But then…

    Out of 138, 89 are contributing at least $5,000 to their yearly income. [ 64% ]

    Out of 138, 71 are contributing at least $10,000 to their yearly income. [ 51% ]

    Out of 138, 47 are contributing at least $20,000 to their yearly income. [ 34% ]

    Out of 138, 31 are contributing at least $35,000 to their yearly income. [ 22% ]

    Out of 138, 26 make $50,000 in income. [ 19% ]

    Out of 138, 18 make $75,000 in income. [ 13% ]

    Out of 138, 16 make $100,000 in income. [ 12% ]

    Out of 138, 10 make $135,000 in income. [ 7% ]

    Out of 138, 6 make $200,000 in income. [ 4% ]

    Out of 138, 4 make $300,000 in income. [ 3% ]

    Out of 138, 1 makes $400,000 in income. [ .72% ]

    Out of 138, 1 makes $500,000 in income. [ .72% ]

    If these numbers are correct, for someone who is working on finishing his first novel and plans to stick with it for the long haul, these percentages are incredibly hopeful.


    1. The results of the AE surveys very closely match what I see in the trenches. It’s all hopeful if you’re willing to write a lot over a five year period.

  9. […] Latest Author Earnings Report | Hugh Howey […]

  10. Just some more evidence that self-published authors are gaining some marketshare in non-fiction:

  11. great article! I would like to reccomend a nice geeky site

Leave a Reply

Your email address will not be published. Required fields are marked *