Author Topic: Latest on sales/borrows to rank?  (Read 2510 times)  

Online MelanieCellier

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 625
  • Gender: Female
  • Australia
    • View Profile
Latest on sales/borrows to rank?
« on: January 19, 2018, 07:33:39 PM »
I can't remember seeing any recent sales/borrows to rank charts. Are there any floating around?

Melanie Cellier | Website | Facebook


Offline PhoenixS

  • Status: A A Milne
  • ******
  • Posts: 4265
  • Gender: Female
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #2 on: January 20, 2018, 05:59:19 AM »
This is about 20 months old now, but spot-checking seems to bear out it's still in the ballpark. At least more so than some of the other calculators out there, many of which haven't been updated since 2012 or 2013. I personally haven't seen anything more updated, and since our catalog is all in with Select, I can't personally pull a new one together myself (because of the ghost borrow effect). Note that this is from the Author Earnings extrapolations, and what Data Guy uses.


Online MelanieCellier

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 625
  • Gender: Female
  • Australia
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #3 on: January 20, 2018, 06:42:51 AM »
This is about 20 months old now, but spot-checking seems to bear out it's still in the ballpark. At least more so than some of the other calculators out there, many of which haven't been updated since 2012 or 2013. I personally haven't seen anything more updated, and since our catalog is all in with Select, I can't personally pull a new one together myself (because of the ghost borrow effect). Note that this is from the Author Earnings extrapolations, and what Data Guy uses.



This is the one I use, good to know it's still vaguely accurate. I just wish it had more granularity! (I'm trying to judge the effectiveness of an AMS ad that's given a sustained boost in ranking without much of a difference in sales, so I'm assuming borrows.)

Melanie Cellier | Website | Facebook

Offline Bill Hiatt

  • Status: Dostoevsky
  • ******
  • Posts: 3323
  • Gender: Male
  • California
    • View Profile
    • Bill Hiatt's Author Website
Re: Latest on sales/borrows to rank?
« Reply #4 on: January 20, 2018, 06:52:14 AM »
This is the one I use, good to know it's still vaguely accurate. I just wish it had more granularity! (I'm trying to judge the effectiveness of an AMS ad that's given a sustained boost in ranking without much of a difference in sales, so I'm assuming borrows.)
Of course, if Amazon would just tell us the number of borrows, we wouldn't have to extrapolate. I know we're not paid on that basis, but I'd still like to have a number. Not only is it good for judging the effectiveness of ads, but it's also a way to help verify audience size. I use "full read equivalents," which is obviously only an approximation.


Tickling the imagination one book at a time
Bill Hiatt | fiction website | education website | Facebook author page | Twitter

Online MelanieCellier

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 625
  • Gender: Female
  • Australia
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #5 on: January 20, 2018, 08:12:47 AM »
Of course, if Amazon would just tell us the number of borrows, we wouldn't have to extrapolate. I know we're not paid on that basis, but I'd still like to have a number. Not only is it good for judging the effectiveness of ads, but it's also a way to help verify audience size. I use "full read equivalents," which is obviously only an approximation.

We can only dream! At this stage I'd settle for any sort of reasonable reporting from AMS at all.

Melanie Cellier | Website | Facebook

Offline PhoenixS

  • Status: A A Milne
  • ******
  • Posts: 4265
  • Gender: Female
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #6 on: January 20, 2018, 08:16:21 AM »
This is the one I use, good to know it's still vaguely accurate. I just wish it had more granularity! (I'm trying to judge the effectiveness of an AMS ad that's given a sustained boost in ranking without much of a difference in sales, so I'm assuming borrows.)

It's simply a predictable mathematical curve, so here's an expanded version I did just for you :). I rounded up, DG rounded down to keep with whole numbers, so there are some 1-digit discrepancies between the two. Wish AMS was so easy to track out!

Now, I believe the sales to maintain also has another history factor than the simple 50% 24-hour decay rate presented here. My hypothesis is that the 50% rate is hit 30 days after consistently being at a rank. I usually depict it as such:
4 days = 10%
1 week = 20%
2 weeks - 30%
3 weeks = 40%
4 weeks = 50%


Offline Not Lu

  • Status: Jane Austen
  • ***
  • Posts: 345
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #7 on: January 20, 2018, 10:28:09 AM »
Here's a list I cobbled together from various sources (Phoenix's list included) plus my own experience. It fills in some of the gaps in Phoenix's list for us little guys. It only has the consistent sales column (average sales from the last 10 days) and assumes you didn't have a rank spike to get to that position.

Rank       Consistent Sales
1          4885
5          2588
20        1301
35        950
100      497
200      311
350      208
500      160
750      117
1000    93
2000    52
3000    37
4000    28
5000    23
6000    20
7000    17
8000    15
9000    13
10000   12
11000   11.5
12000   11
13000   10
14000   9
15000   8
16000   7
20000   6
25000   5
30000   4
40000   3
50000   2.5
100000   1
« Last Edit: January 20, 2018, 10:30:06 AM by Not Lu »

Offline Decon

  • Status: A A Milne
  • ******
  • Posts: 4517
    • View Profile
    • Declan Conner, Blog
Re: Latest on sales/borrows to rank?
« Reply #8 on: January 21, 2018, 04:47:22 AM »
Here's a list I cobbled together from various sources (Phoenix's list included) plus my own experience. It fills in some of the gaps in Phoenix's list for us little guys. It only has the consistent sales column (average sales from the last 10 days) and assumes you didn't have a rank spike to get to that position.

Rank       Consistent Sales
1          4885
5          2588
20        1301
35        950
100      497
200      311
350      208
500      160
750      117
1000    93
2000    52
3000    37
4000    28
5000    23
6000    20
7000    17
8000    15
9000    13
10000   12
11000   11.5
12000   11
13000   10
14000   9
15000   8
16000   7
20000   6
25000   5
30000   4
40000   3
50000   2.5
100000   1


Id say that was reasonably accurate at the lower end. I've been averaging around 6 inclusive sales/borrows per day and hovered around the 20.000 mark during an AMS promo that finished a few days ago.

While the main chart is important, it's the cat ranks that that are the most important for ongoing sales. At a rank of 20,000 in my genre it put me at 126 in the cat ranks which is not good enough to create visibility on its own.


Scorpion ebooks: Full length  thrillers with a sting in the tale. All enrolled in KU & Prime.
Declan Conner | blog

Online TellNotShow

  • Status: Jane Austen
  • ***
  • Posts: 305
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #9 on: January 22, 2018, 04:54:27 PM »
I'm confused. Do these charts represent the approximate number of actual sales required for books in KU to achieve a certain ranking?

Or is there some allowance being made for borrows, built into the number in some way, for instance a full read through, a half read through or whatever?
Descent Into Madness: 99%

How To Win KBoards Friends and Not Puss People Off: 0.1%

Writing Desk Space Won Back From Dangerous Cat: 0.2%
All things in moderation

Online Amanda M. Lee

  • Status: Edgar Allan Poe
  • *******
  • Posts: 5170
  • Gender: Female
  • Listen to me, you will.
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #10 on: January 22, 2018, 05:03:59 PM »
I'm confused. Do these charts represent the approximate number of actual sales required for books in KU to achieve a certain ranking?

Or is there some allowance being made for borrows, built into the number in some way, for instance a full read through, a half read through or whatever?
A borrow gets the equivalent rank of one sale no matter if its ever opened. Page reads dont change rank at all.

Amanda M. Lee

Online MelanieCellier

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 625
  • Gender: Female
  • Australia
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #11 on: January 22, 2018, 06:01:14 PM »
It's simply a predictable mathematical curve, so here's an expanded version I did just for you :). I rounded up, DG rounded down to keep with whole numbers, so there are some 1-digit discrepancies between the two. Wish AMS was so easy to track out!

Now, I believe the sales to maintain also has another history factor than the simple 50% 24-hour decay rate presented here. My hypothesis is that the 50% rate is hit 30 days after consistently being at a rank. I usually depict it as such:
4 days = 10%
1 week = 20%
2 weeks - 30%
3 weeks = 40%
4 weeks = 50%



Thanks Phoenix! Math is not my forte ;D

I must be getting a lot of ghost borrows because I'm definitely not getting the page reads I'd expect for what my borrow numbers must be. But then, this is my first box set, so maybe it's normal for them to perform differently than individual books.

Melanie Cellier | Website | Facebook

Offline thevoiceofone

  • Status: Arthur C Clarke
  • *****
  • Posts: 2043
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #12 on: January 23, 2018, 08:25:01 AM »
nice chart.

Online TellNotShow

  • Status: Jane Austen
  • ***
  • Posts: 305
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #13 on: January 23, 2018, 06:26:24 PM »
A borrow gets the equivalent rank of one sale no matter if its ever opened. Page reads dont change rank at all.

Yes, I understand that. My question was about what these charts actually mean, rather than how rank is calculated.

Perhaps I wasn't clear enough with my question, I'll try again.

Are these charts rank estimates for KU only books, and the sales numbers actual sales numbers? I'm wondering if people are including some sort of estimated borrows WITHIN the quoted sales figures, or they're ACTUAL sales, then there are borrows AS WELL. (Sorry for the caps, just an emphasis for clarity, I'm not shouting.)

For instance, Decon's comment said "Id say that was reasonably accurate at the lower end. I've been averaging around 6 inclusive sales/borrows per day and hovered around the 20.000 mark during an AMS promo that finished a few days ago."
Which indicates that, perhaps Decon was getting, say, 3 sales a day and assuming three borrows because of an estimate based on page reads.

Hope that's clear enough now. Just trying to work out for myself whether it's worth it to me to put something into KU, and these charts would help if I was sure what they actually represented.
Descent Into Madness: 99%

How To Win KBoards Friends and Not Puss People Off: 0.1%

Writing Desk Space Won Back From Dangerous Cat: 0.2%
All things in moderation

Online MelanieCellier

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 625
  • Gender: Female
  • Australia
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #14 on: January 24, 2018, 02:09:44 AM »
Yes, I understand that. My question was about what these charts actually mean, rather than how rank is calculated.

Perhaps I wasn't clear enough with my question, I'll try again.

Are these charts rank estimates for KU only books, and the sales numbers actual sales numbers? I'm wondering if people are including some sort of estimated borrows WITHIN the quoted sales figures, or they're ACTUAL sales, then there are borrows AS WELL. (Sorry for the caps, just an emphasis for clarity, I'm not shouting.)

For instance, Decon's comment said "Id say that was reasonably accurate at the lower end. I've been averaging around 6 inclusive sales/borrows per day and hovered around the 20.000 mark during an AMS promo that finished a few days ago."
Which indicates that, perhaps Decon was getting, say, 3 sales a day and assuming three borrows because of an estimate based on page reads.

Hope that's clear enough now. Just trying to work out for myself whether it's worth it to me to put something into KU, and these charts would help if I was sure what they actually represented.

My understanding is that these charts are based on sales for non-KU books where there is a clear sales to rank correlation.

Those of us in KU are using it as a rough guide for how many borrows we're getting at any given time. (i.e. My rank suggests I should have around 20 sales today from .com, I have 10 sales, therefore, I must have had roughly 10 borrows. Calculating the number of page reads will then give you a rough idea of 'ghost borrows' which are borrows that don't result in page reads. But of course it all starts to get very hazy at that point since it might be five full read throughs and five ghost borrows, or ten borrows that resulted in only partial read throughs, etc. Plus you don't know how long subscribers will hold onto a borrowed book before starting it, so it all becomes very rough approximations at that point.)

Melanie Cellier | Website | Facebook

Online MelanieCellier

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 625
  • Gender: Female
  • Australia
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #15 on: January 24, 2018, 02:11:32 AM »
Someone asked Data Guy about it in the comments on the latest Author Earnings report, and it sounds like he won't be releasing charts like this publicly anymore. Pity. I wonder which industry players don't want us having this information.

Melanie Cellier | Website | Facebook

Offline Michael_J_Sullivan

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 598
  • Gender: Male
  • Fairfax, VA
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #16 on: January 24, 2018, 06:13:00 AM »
Does anyone know of sales rank to sales ratios that are adjusted for various months?  Or that get adjusted over the years?  To hit 1,000 in 2016 is much different than hitting 1,000 in 2018. Likewise 1,000 ranking in April is much different than sales at that same ranking in December.
Wikipedia | Amazon | Twitter | Blog/Website |  Bundled Hollow World DRM-free ebooks with purchase of print or audio

Offline KevinH

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 825
  • Gender: Male
    • View Profile
    • The Hardman Writing Stylus
Re: Latest on sales/borrows to rank?
« Reply #17 on: January 24, 2018, 07:11:51 AM »
I've got a page linking to a number of Sales Rank calculators and such on my blog:  https://kevinhardman.blogspot.com/p/kindle-sales-rank.html

When I was wide I thought they were generally on the money.  Being in KU, I think they're useful - as someone noted - for estimating borrows.

Offline It's A Mystery

  • Status: Jane Austen
  • ***
  • Posts: 350
  • Oxford
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #18 on: February 13, 2018, 06:50:50 AM »
Here's a list I cobbled together from various sources (Phoenix's list included) plus my own experience. It fills in some of the gaps in Phoenix's list for us little guys. It only has the consistent sales column (average sales from the last 10 days) and assumes you didn't have a rank spike to get to that position.

Rank       Consistent Sales
1          4885
5          2588
20        1301
35        950
100      497
200      311
350      208
500      160
750      117
1000    93
2000    52
3000    37
4000    28
5000    23
6000    20
7000    17
8000    15
9000    13
10000   12
11000   11.5
12000   11
13000   10
14000   9
15000   8
16000   7
20000   6
25000   5
30000   4
40000   3
50000   2.5
100000   1

Thanks for this (and Pheonix for hers), but this is making me doubt whether sales and borrows are in fact treated equally.

My last four sales days for a single book on Amazon US store are 6, 12, 10 and 17 with today having 6 so far. My rank is currently 16,588 (the highest it's been), which doesn't seem to be high enough against the above chart. I'm not in select so these are all sales.

Offline Cassie Leigh

  • Status: Scheherazade
  • *****
  • Posts: 1829
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #19 on: February 13, 2018, 08:08:27 AM »
You aren't holding steady at a rank, so you'd need to use the left-hand side of that chart.
6 sales=45,000 or so
12 sales =35,000 or so
10 sales = 40,000 or so
17 sales = 20,000 or so

You're probably in the 17,000 range because you have upward momentum that's being taken into account. But until you have a steady level of sales for a couple of weeks, the right-hand side of the chart won't apply to you.


8 Pen Names. Genres: Non-fiction, Speculative Fiction, Romance.
Blog | Video Courses | Quick & Dirty Guide to AMS

Offline It's A Mystery

  • Status: Jane Austen
  • ***
  • Posts: 350
  • Oxford
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #20 on: February 13, 2018, 08:11:53 AM »
You aren't holding steady at a rank, so you'd need to use the left-hand side of that chart.
6 sales=45,000 or so
12 sales =35,000 or so
10 sales = 40,000 or so
17 sales = 20,000 or so

You're probably in the 17,000 range because you have upward momentum that's being taken into account. But until you have a steady level of sales for a couple of weeks, the right-hand side of the chart won't apply to you.

Ah I see, the yellow chart makes more sense!

The five days before the ones listed were 7, 8, 6, 9 and 9 so how long do these take to 'stick' as it were?

Offline PhoenixS

  • Status: A A Milne
  • ******
  • Posts: 4265
  • Gender: Female
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #21 on: February 13, 2018, 08:53:26 AM »
Ah I see, the yellow chart makes more sense!

The five days before the ones listed were 7, 8, 6, 9 and 9 so how long do these take to 'stick' as it were?

That's where this bit above figures in:
Quote
Now, I believe the sales to maintain also has another history factor than the simple 50% 24-hour decay rate presented here. My hypothesis is that the 50% rate is hit 30 days after consistently being at a rank. I usually depict it as such:
4 days = 10%
1 week = 20%
2 weeks - 30%
3 weeks = 40%
4 weeks = 50%

This is, of course, simply ballpark. There is also a small history factor that plays into it (and new releases have no history...unless they had preorder sales). And there is a velocity factor. So, depending on when you're looking at your rank, it could be sales late in the day from the day before haven't dropped out yet. That means if you had, for example, 30 sales yesterday and 10 of those happened after 1:00pm and you've had 2 sales today before 1:00pm, and you're looking at your rank at 5:00pm today (because there's about a 4-6 hour lag in rank updates when a book is selling), then your rank would reflect the 10 sales from yesterday (that happened within the last 24-hour period) plus the 2 sales today.

And to Michael Sullivan's point above, time of year makes a difference as store sales velocity figures into this too because it's a relational algo -- your book is being ranked in relation to the sales of other books in the store rather than to an absolute. February historically has lower ebook sales than, say, late December and early January, when store velocity can be up by 15-25%.

And as a book nears the #1000 rank, even the big Sunday KDDs and Gold Box deals that can have up to 50 or 100 books can influence those lower/better ranks.

So, lots of moving parts, and lots to take into account when trying to figure out how to correlate rank to sales/borrows. The above charts are simply ballpark estimates and averages.

Offline It's A Mystery

  • Status: Jane Austen
  • ***
  • Posts: 350
  • Oxford
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #22 on: February 13, 2018, 09:00:36 AM »
That's where this bit above figures in:
This is, of course, simply ballpark. There is also a small history factor that plays into it (and new releases have no history...unless they had preorder sales). And there is a velocity factor. So, depending on when you're looking at your rank, it could be sales late in the day from the day before haven't dropped out yet. That means if you had, for example, 30 sales yesterday and 10 of those happened after 1:00pm and you've had 2 sales today before 1:00pm, and you're looking at your rank at 5:00pm today (because there's about a 4-6 hour lag in rank updates when a book is selling), then your rank would reflect the 10 sales from yesterday (that happened within the last 24-hour period) plus the 2 sales today.

And to Michael Sullivan's point above, time of year makes a difference as store sales velocity figures into this too because it's a relational algo -- your book is being ranked in relation to the sales of other books in the store rather than to an absolute. February historically has lower ebook sales than, say, late December and early January, when store velocity can be up by 15-25%.

And as a book nears the #1000 rank, even the big Sunday KDDs and Gold Box deals that can have up to 50 or 100 books can influence those lower/better ranks.

So, lots of moving parts, and lots to take into account when trying to figure out how to correlate rank to sales/borrows. The above charts are simply ballpark estimates and averages.

Brilliant, thanks Pheonix. Makes total sense now.


Offline PhoenixS

  • Status: A A Milne
  • ******
  • Posts: 4265
  • Gender: Female
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #23 on: February 13, 2018, 09:04:22 AM »
Someone asked Data Guy about it in the comments on the latest Author Earnings report, and it sounds like he won't be releasing charts like this publicly anymore. Pity. I wonder which industry players don't want us having this information.

That would be a real shame, since I believe I was the one who pointed out the maintenance factor and helped demonstrate how his initial estimates of sales was too high and how to correct for it. Lots of indies have offered up information that helped him build his modeling -- and we did it for free. And offered it to the community for free. :(

But you know, we figured out the model long before he refined it with his spiders and heavy processing power. We're resourceful -- I'm betting we can do our own modeling once again. ;)

Online Dpock

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 681
  • Gender: Male
  • North Idaho
    • View Profile
Re: Latest on sales/borrows to rank?
« Reply #24 on: February 13, 2018, 09:41:55 AM »
(because there's about a 4-6 hour lag in rank updates when a book is selling)


Is that the gospel truth? Books are always selling. My understanding, which may be flawed, is if a book is ranking at 50k at noon, then 40k at 1:00 pm, it means a sale or borrow occurred "roughly" in the interim.