Update to the Fantasy Bestseller Breakdown

20 Comments / By /

Back in February, I posted the results of some data mining, specifically about the bestseller lists for fantasy and science fiction. You can see the old article here.
It seemed like now might be a good time for an update. There\’s been much to-do about the change to the Amazon algorithms. The very-important \”Popularity\” ranking has seen significant changes in March and again in May. The new algorithms have made major changes to how books show up in the \”Popular\” ranking – which is the default manner customers see ebooks on Amazon, making it vital to sales.
Roughly how it works: you get sales, your rank goes up. The more sales you get over the period tracked, compared to other books, the more your rank goes up. The old system made loans via the Select program about the same as sales, and gave a lot of weight to free books given away through Select promotion periods. The new system rates free books at perhaps 10% of the value of sold books, does not seem to give any boost for loaned books, and most crucial – it seems to factor price into the equation, weighting higher priced books more heavily.
So how does this affect indies, whose books have been selling like hotcakes because they\’re priced lower? If Amazon weights higher priced books higher, then this makes breaking out at $0.99 or $2.99 that much more difficult.
It\’s the end of June, four months since my last survey and over a month since the May algorithm changes. Seemed like it was time for new data!
 

Fantasy Genre, Top Hundred Bestsellers

Well, top 95 bestsellers, anyway. There seem to be some issues with doubled up versions of some of George R.R. Martin\’s books, which resulted in only 95 books actually listed in the top 100.
The breakdown was 53 (56%) traditionally published books vs 42 (44%) self published books. This is a significant change. In fact, it\’s the first time this year that I\’ve seen the genre drop much below 50% indie/self pub books; in February, indie books were 53% of the top list for fantasy, and it\’s stayed at that level +/- about 4% through early May.
Some other interesting data points:
Indie price average: $3.24
Trad price average: $9.56
Overall average price: $6.77
Breakdown follows. Note, there were several books I raised to the X.99 levels to make the data easier to chart. These were Indie at $2.51 and $4.95, and Trad at $4.90, $7.29, $7.39, $8.32, and $9.34 (actual numbers were used for the averages above, no rounding).
$0.99 – Indie 6 (down 7), Trad 0 (same), Total 6 (6%)
$1.99 – Indie 2 (same), Trad 0 (same), Total 2 (2%)
$2.99 – Indie 16 (down 7), Trad 0, Total 16 (17%)
$3.99 – Indie 13 (up 1), Trad 0, Total 13 (14%)
$4.99 – Indie 4 (up 2), Trad 2 (up 1), Total 6 (6%)
$5.99 – Indie 0, Trad 0, Total 0 (0%)
$6.99 – Indie 1 (up 1), Trad 0 (down 1), Total 1 (1%)
$7.99 – Indie 0 (same), Trad 20 (same), Total 20 (21%)
$8.99 – Indie 0 (same), Trad 15 (up 4), Total 15 (16%)
$9.99 – Indie 0 (same), Trad 7 (up 4), Total 7 (7%)
$10.99 Total 0 (same)
$11.99 Indie 0 (same), Trad 1 (same), Total 1 (1%)
$12.99 Indie 0 (same), Trad 6 (same), Total 6 (6%)
$13.99 Total 0 (same) (0%)
$14.99 Indie 0 (same), Trad 1 (down 2), Total 1 (1%)
$29.99 Indie 0 (same), Trad 1 (same), Total 1 (1%) (Martin boxed set)
 

Fantasy Graph

As ever, a picture is worth a thousand words:
\"\"
 

Thoughts?

This is not enough data to draw conclusions from. However, coupled with the excellent analysis done by Ed Robertson, it\’s possible to hypothesize that the changes he noted to the Amazon book algorithms are damaging indie sales penetration. Some of this might be the loss of impact from Kindle Select promotions and loans. Some could also be due to algorithms favoring higher prices (note the upward trend across the board for prices of indie books which made the top ranks).
I think we need to do substantial more work on data crunching to see where things are headed, and I\’m open to collaboration with others. In the short term, however, I think it\’s extremely likely many indies are selling themselves short by pricing too low – to their detriment.
Hopefully this helps some folks out.
 
Thanks for stopping by! I hope you found this information interesting and useful. I\’d love to hear your thoughts on the data, what it might mean, and how writers can best adapt to these changes.
To celebrate a year of indie publishing, I have my novel, \”By Darkness Revealed\” reduced to FREE on Amazon for today and tomorrow – 6/21 + 6/22 only. If you enjoy urban fantasy, you might enjoy checking it out, and you\’ll aid my in my quest to break the top 100 free books! http://www.amazon.com/Darkness-Revealed-Blackwell-Magic-ebook/dp/B005G8L3X4/ref=sr_1_1?ie=UTF8&qid=1340283711&sr=8-1&keywords=by+darkness+revealed

20 thoughts on “Update to the Fantasy Bestseller Breakdown”

  3. M. Louisa Locke

    Thanks a lot, this pretty much confirms what I have been seeing in the historical mystery category.
    I too believe this is both a factor of the change in the way that free promotions and borrows are counted so that they are no longer giving indie books a distinct advantage in terms of rankings, and the change in the impact of price.
    What I would love to know is if it is actual price or revenue which is now a factor, and over what period of time (total sales as ebooks over the life of the book, or the past 30 days, the past hour?) I did raise my price last month to $3.99, so I seem to be right on target in terms of the trends in pricing, but I certainly don’t think I will ever feel comfortable raising my prices much higher–because of my readers. (That is why I am interested in if it is just price or revenue. If the higher prices books don’t sell much, their total revenue would be less than a lower priced book ($2.99 and above) that sold more. I would think that Amazon would be more interested in how much a book was making them versus the simple price.
    What I am seeing is the effect of the purported 30 days sales factor for popularity that Ed talked about. I did another KDP free promotion for my first book , Maids of Misfortune, a month ago, and did my sequel, Uneasy Spirits, a week later. They both stayed very firmly in the top 10 of the historical mystery popularity category after the promotion, but slowly slipped in the best seller list (and tend to bounce around as they go down). But suddenly the first book has started to slip on the popularity list as well (but not the second–even though the first has consistently been selling better, and been higher in the best seller rankings.)
    My conclusion is that Maids has just passed the 30 day mark, and the sales bump of the promotion has dropped out of the equation, which means the newer, lower sales days now are having a greater effect. i suspect that in a few days the second book, passing the 30 day mark, will start to slip as well.
    Interesting stuff that is definitely going to keep me on my toes.
    Anyway, thanks again for the update.
    M. Louisa Locke
    Maids of Misfortune and Uneasy Spirits

    Reply
    1. Kevin

      I agree, overall. What we’re seeing, I think, is some level of revenue based placement in the popularity listing. So a 99 cent book needs to sell significantly more copies than a $10 ebook to stay in the most popular ranks.
      And popularity, because it is the default parameter for people browsing for books on Amazon (and because most people never change the default on anything) is a prime mover for bestseller ranking. Placement on the popularity index sells books.
      Your assessment of your own books looks dead on to me. As for pricing, I suggest watching carefully where Amazon places their own books in a given genre. They are the ones with the entire suite of data in front of them. They know consumer habits, and they know their own algorithms. Odds are good that they’re pricing in a very scientific manner based on better data than anyone else in the business has available.

      Reply
  4. M. Louisa Locke

    Thanks a lot, this pretty much confirms what I have been seeing in the historical mystery category.
    I too believe this is both a factor of the change in the way that free promotions and borrows are counted so that they are no longer giving indie books a distinct advantage in terms of rankings, and the change in the impact of price.

    What I would love to know is if it is actual price or revenue which is now a factor, and over what period of time (total sales as ebooks over the life of the book, or the past 30 days, the past hour?) I did raise my price last month to $3.99, so I seem to be right on target in terms of the trends in pricing, but I certainly don’t think I will ever feel comfortable raising my prices much higher–because of my readers. (That is why I am interested in if it is just price or revenue. If the higher prices books don’t sell much, their total revenue would be less than a lower priced book ($2.99 and above) that sold more. I would think that Amazon would be more interested in how much a book was making them versus the simple price.

    What I am seeing is the effect of the purported 30 days sales factor for popularity that Ed talked about. I did another KDP free promotion for my first book , Maids of Misfortune, a month ago, and did my sequel, Uneasy Spirits, a week later. They both stayed very firmly in the top 10 of the historical mystery popularity category after the promotion, but slowly slipped in the best seller list (and tend to bounce around as they go down). But suddenly the first book has started to slip on the popularity list as well (but not the second–even though the first has consistently been selling better, and been higher in the best seller rankings.)

    My conclusion is that Maids has just passed the 30 day mark, and the sales bump of the promotion has dropped out of the equation, which means the newer, lower sales days now are having a greater effect. i suspect that in a few days the second book, passing the 30 day mark, will start to slip as well.

    Interesting stuff that is definitely going to keep me on my toes.
    Anyway, thanks again for the update.

    M. Louisa Locke
    Maids of Misfortune and Uneasy Spirits

    Reply
    1. Kevin

      I agree, overall. What we’re seeing, I think, is some level of revenue based placement in the popularity listing. So a 99 cent book needs to sell significantly more copies than a $10 ebook to stay in the most popular ranks.
      And popularity, because it is the default parameter for people browsing for books on Amazon (and because most people never change the default on anything) is a prime mover for bestseller ranking. Placement on the popularity index sells books.

      Your assessment of your own books looks dead on to me. As for pricing, I suggest watching carefully where Amazon places their own books in a given genre. They are the ones with the entire suite of data in front of them. They know consumer habits, and they know their own algorithms. Odds are good that they’re pricing in a very scientific manner based on better data than anyone else in the business has available.

      Reply
  5. Merry Bond

    Wow, what fabulous information. I wish I knew how I could use it to make my book show up on those popularity lists. Magic In The Storm has just fallen off it’s KDP Select promotion bump and now I’m back to the piddly sales I had before.
    But at least now I know that I’m pricing it right at 2.99!
    I look forward to seeing your romance stats since MITS is both fantasy and romance. 🙂

    Reply
  6. Merry Bond

    Wow, what fabulous information. I wish I knew how I could use it to make my book show up on those popularity lists. Magic In The Storm has just fallen off it’s KDP Select promotion bump and now I’m back to the piddly sales I had before. But at least now I know that I’m pricing it right at 2.99!
    I look forward to seeing your romance stats since MITS is both fantasy and romance. 🙂

    Reply

  7. Pingback: More From DWS on Pricing « Tales From The Mad Monk

  9. Pingback: Amazon: Pricey Best Sellers Breakdown | Adventures in Text

  11. Charl Dee

    Good work. I posted this on Ed’s Part 2, also.
    If you want to catch snipe, you have to think like a snipe.
    Consider; all the ereaders record what sentences are highlighted, leading to reports of most highlighted. Therefore, they (Amazon, B&N, Kobo, etc.) know when and how much of a book has been read. This will be true whether it was sold, borrowed or given away free. The act of reading all of a book is probably the most reliable data point for popularity available. No doubt it took some time to accumulate statisticallyly significant data on the Select program. However, these several changes are probably driven by better data. I’m not trying to sell books, so I’ll have to leave the consideration of the impact to others such as you.

    Reply
  12. Charl Dee

    Good work. I posted this on Ed’s Part 2, also.If you want to catch snipe, you have to think like a snipe.
    Consider; all the ereaders record what sentences are highlighted, leading to reports of most highlighted. Therefore, they (Amazon, B&N, Kobo, etc.) know when and how much of a book has been read. This will be true whether it was sold, borrowed or given away free. The act of reading all of a book is probably the most reliable data point for popularity available. No doubt it took some time to accumulate statisticallyly significant data on the Select program. However, these several changes are probably driven by better data. I’m not trying to sell books, so I’ll have to leave the consideration of the impact to others such as you.

    Reply

  15. Pingback: Honest opinions on my price? - Page 3

Leave a Comment

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

Scroll to Top