Hugh Howey presents: Amazon author earnings

[banjomike]

"But the staggering number of reviews involved for most of these books (over a hundred on average across our entire sample) makes each of these highly unlikely. "

I'm guessing that someone has not seen the websites and discussion sites where authors (and their 'helpers') swap reviews in bulk or the sites that openly advertise the sale of bulk reviews. "Me and my dozen users will review you, and you and your dozen can review me", "Pay me and I will add twenty reviews to your book". Sometimes it is more obvious than at other times but it certainly happens on Goodreads where there is considerable enforcement against it happening. Amazon is not going to be immune to it.

Apart from that, I loved the article.

[fjtorres]

The Bookseller talked to Howey about the website and the ongoing effort:
http://www.thebookseller.com/news/ho...ysis-site.html

An upcoming report, he says, will look at 50,000 titles “across all genres.” In the interest of transparency, the report also offers an Excel spreadsheet with base data and graphics.
The Author Earnings site includes not only the initial report on these sales-ranking analyses in three major genres but also a short survey for users to take and a petition in which users are asked to rank their interest in various calls for action. The statements for which users signal their levels of support on the petition include: “Authors deserve, at minimum, 50% of net sales on e-books”; “All publishing contracts should be for a limited term of license (5 to 10 years)”; “DRM should be abolished for all e-books”; and a call for “Complete abolishment of non-compete clauses” in author contracts.

[hardcastle]

Originally Posted by fjtorres

Actually, on any given day some subcategories can be headed with single digit sales. Others might need thousands. Amazon slices and dices their categories to help buyers find books so there are lots of categories to slot books in.

This data set is a time slice snapshot.
Further slices will provide further snapshots to enable time-based analysis.

As far as I can tell, most of the report focuses on a single category: what the author deems as "genre" fiction, or an amalgamation of sci-fi, romance, mystery, fantasy, and the like. I should dive into the data myself to see how this is categorized, because as good as the write-up was, it doesn't address some very basic aspects and could be ignoring thousands of relevant data points.

I was not talking about variability among micro-categories or even time, though they are also extremely relevant.

[BearMountainBooks]

Originally Posted by fjtorres

Shrug.
Nobody forces you to believe anything.
But the study includes the raw data and it will be analyzed to death in the days to come.
Those open to numerical analysis will consider the data; those that aren't will keep on denying. Doesn't matter.
In the end, we as readers aren't the target audience.
It is midlist authors and newcomers.
It will only affect us over time as authors start going indie more often or use the data to get the publishers to actually negotiate living wage royalties so they can keep on writing.

It was also pointed out in the article that it's possible people are more lenient in their reviews if they paid less. I'm inclined to believe it because I know I do it. Not really on purpose, but I noticed my own behavior a while back. If I pay 8 dollars for a book and it stinks...I'm MAD about the wasted time AND money. My review tends to reflect that anger. If it cost 99 cents, I might not even be mad enough to review it at all. I used to review everything I read. Nowadays I can't be bothered to review them all. It depends on time available, mood and whether or not I have anything interesting to say (this does not include review copies. I tend to review ARCS because I agreed to review them. Sometimes I am annoyed at having to review an ARC, especially if the book was a stinker. This annoyance may also show up in the review.)

My behavior changes a bit if there is a hot button issue in a book. Certain sloppiness in plotting or certain plots annoy the bajeebers out of me. I'm going to be quicker to take stars away. Example: If you have a great UF that suddenly turns into a "I am wildly attracted to anything and everything that moves" at any point during the book, the stars are going to fall so fast there won't be any left...

Of course, I am but one reviewer. Not everyone reviews like I do, but it's quite possible that the data supports a price-affects-the-review.

That said, I think there are a number of liberties taken in analyzing the data...

[peachiekene]

Originally Posted by BearMountainBooks

Of course, I am but one reviewer. Not everyone reviews like I do, but it's quite possible that the data supports a price-affects-the-review.

That said, I think there are a number of liberties taken in analyzing the data...

I tend to agree with the price affecting the review. I've noticed that in the past with myself.
It can also depend on the type of book. If I picked it up because I thought it "looked cute" or something - I'm more lenient because I wasn't really expecting much from it anyway. (I no longer buy books without skimming the first chapter anymore because there are so many poorly edited books out there - I've been burned too many times.)

I'm wondering where the ratings are coming from. If you go to any well reviewed book on goodreads, they all average between 3-4.25 stars, not 4.25 - 4.5 like the data suggests.
Books that have less reviews have higher average scores. Even generally beloved books don't break the 4.25.
-Pride and Prejudice: 4.24
-Hitchhiker's Guide to the Galaxy: 4.17
-Gone with the Wind: 4.24

The exceptions would be books in a series/omnibuses.

The lower ratings make far more sense because there's no one thing that every single person absolutely loves.

Not that goodreads has the authority on reviews - I just don't think the place people are buying the books from is necessarily the best source.

I've also noticed more recently published books (especially YA and NA) tend to have much higher ratings - and I wonder if this is because of the few scandals involving authors or author-fans harassing negative reviews - as well as people leaving goodreads.

[BearMountainBooks]

Originally Posted by peachiekene

I tend to agree with the price affecting the review. I've noticed that in the past with myself.
It can also depend on the type of book. If I picked it up because I thought it "looked cute" or something - I'm more lenient because I wasn't really expecting much from it anyway. (I no longer buy books without skimming the first chapter anymore because there are so many poorly edited books out there - I've been burned too many times.)

I'm wondering where the ratings are coming from. If you go to any well reviewed book on goodreads, they all average between 3-4.25 stars, not 4.25 - 4.5 like the data suggests.
Books that have less reviews have higher average scores. Even generally beloved books don't break the 4.25.
-Pride and Prejudice: 4.24
-Hitchhiker's Guide to the Galaxy: 4.17
-Gone with the Wind: 4.24

The exceptions would be books in a series/omnibuses.

The lower ratings make far more sense because there's no one thing that every single person absolutely loves.

Not that goodreads has the authority on reviews - I just don't think the place people are buying the books from is necessarily the best source.

I've also noticed more recently published books (especially YA and NA) tend to have much higher ratings - and I wonder if this is because of the few scandals involving authors or author-fans harassing negative reviews - as well as people leaving goodreads.

Some speculation: GR encourages two star reviews as "I liked it." Amazon lists two stars as "I didn't like it." Not everyone bothers to look at what the stars supposedly mean but where I will sometimes give a 2 star on GR because it supposedly means "I liked it" I will usually bump it to 3 on Amazon. But it can be arbitrary.

The averages are probably computed differently too. I notice it most with my own books, but a three star drags the average down more on GR than on Amazon. Amazon wants to sell books--as a retailer, I think they err on the side of presenting a higher average (more of a bell curve maybe?) I've never sat and done the averaging out--note I said this was speculation on my part!

There are have been just as many controversies on Amazon over reviews, traded reviews, etc. Amazon even started taking down reviews if another author was obvious about being an author (apparently with the tit for tat exchange of reviews, some authors thought it was a bright idea to mention their own books or put author in their name and so on.)

Reviews are an odd business.

Another thing that skews reviews is when you get to second and third books, you are generally left with people who LIKED the series. So you GENERALLY start seeing less two and one star reviews. I say generally because one of my series has two star reviews from the same person across the entire series. Now, even knowing that GR means "I liked it" I'm amazed that anyone "liked" it to that extent, but kept reading the entire series. I think that is an exception though. Most people, if a book is 3 or 2 stars are going to be on the fence about continuing the series.

Then too, with series, there can be a stinker book in the middle. OR some reviewers get "harder" on the series, expecting MORE with each book or comparing it to their favorite in the series. I have a hard time not doing that myself. I love Ilona Andrews, John Levitt, etc. When I review their books, I tend to think, "Well the first is the best." So do a penalize the second or third a star because I'm not comparing it to books in general? I find I lean in that direction...

Complicated.

[Yapyap]

Two stars means "It was OK" on Goodreads, according to the hints they display next to the stars. Three stars is "I liked it" (while it's "it was OK" on Amazon).

As for someone reading the entire series and rating every book two stars - could be someone who rates "low" in general, and is perfectly happy to read books they find "OK" (and perhaps already bought the entire series, therefore don't feel any particular desire to drop it as it's not like they're actually disliking it).

I've seen quite a few people who tend to rate low - they'll praise a book in the review and give it three stars, while leaving the fours and fives to books they (perhaps) feel aren't just good entertainment / an enjoyable pastime but true pinnacles of literary achievement.

But yes, subsequent books in series do tend to get higher ratings than the first one, and this makes perfect sense - people who didn't care much for the first one are usually not going to read more, so unless the author writes a real stinker, or the series takes a sharp turn towards something it wasn't to begin with, most people who keep reading will be those who are likely to enjoy it quite a bit.

As for reviews - I also noticed, back when I tried to read at least one self-published book a month, that unless the book was completely horrible / illiterate / completely lacked anything reminiscent of plot, grammar or coherence, I tended to be more lenient with my ratings and reviews compared to trad-pubbed books - I would rate something four stars when a similar-quality trad-pubbed book wouldn't have got above three, and so on.

I'm not even sure why that was - a combination of lower price and actively wanting to be encouraging, maybe? I wanted the not-illiterate self-published authors I read to succeed?

There's definitely some element of rating and reviewing differently. I'm not sure how significant it is, or how many readers even pay attention to whether a book is self-pubbed or trad-pubbed, but it's one factor to consider.

[HarryT]

Is it Goodreads that lets you rate a book before it's even released? I know that one of the main "reading" sites does. That, to me, seems to completely devalue any claim to objectivity that ratings might have.

[BearMountainBooks]

Originally Posted by HarryT

Is it Goodreads that lets you rate a book before it's even released? I know that one of the main "reading" sites does. That, to me, seems to completely devalue any claim to objectivity that ratings might have.

Why? I get ARCs and so do a lot of people. We've read the book and usually say it was an ARC in the review. Netgalley is now open to applications from the general public (although those with blogs seem to get more books). I try to wait until close to the release, but I often have a large pile so I will easily get in a week early...or late!

I thought GR was 2 star "I liked it." Or maybe the hint is different than the actual or maybe they changed it. I know people were asking them to change it because two stars really doesn't convey "I liked it." Anyway, I could be wrong.

I think I rate indie work higher sometimes, but that's "appears" to me to be because of the lower price I paid. There could EASILY be subconscious solidarity there too. But if a book is bad or annoys me, watch out. It won't matter if the book was free...

[fjtorres]

Originally Posted by HarryT

Is it Goodreads that lets you rate a book before it's even released? I know that one of the main "reading" sites does. That, to me, seems to completely devalue any claim to objectivity that ratings might have.

Baen *sells* eARCs.

http://www.baenebooks.com/c-2-advanc...er-copies.aspx

And plenty of established authors seed the market with ARCs. It's actually one of the few marketting tools BPHs use for midlisters.

[HarryT]

Originally Posted by fjtorres

Baen *sells* eARCs.
And plenty of established authors seed the market with ARCs. It's actually one of the few marketting tools BPHs use for midlisters.

I'm not talking about that. I'm talking about the practice of giving a book 5* reviews merely on its announcement, for no better reason than that the reviewer likes the author. There's at least one of the book sites, and I can't remember which one it is - it may be goodreads or it may be a different one - on which this practice is widespread.

[fjtorres]

Originally Posted by HarryT

I'm not talking about that. I'm talking about the practice of giving a book 5* reviews merely on its announcement, for no better reason than that the reviewer likes the author. There's at least one of the book sites, and I can't remember which one it is - it may be goodreads or it may be a different one - on which this practice is widespread.

Worse ones, too.
Once you start crowd sourcing you take the good with the bad.
Doesn't mean the bad makes the good worthless.
All it means is CAVEAT EMPTOR applies there, too.

[fjtorres]

Here's an interview with Howey on the Author Earnings project:
http://www.youtube.com/watch?v=2gaey6a0zgA#t=5367

[Lemurion]

All these numbers are interesting, but I don't know that anyone has enough of the right numbers to make more than simple sweeping generalizations.

These numbers clearly show self-published authors can make a lot of money on Amazon, something we have plenty of other data points to support as well, but I don't know that they do more than that.

It's been clear for a while now that for most authors, self-published success is going to produce more income than equivalent commercially published success. Let's face it, Amazon's 70% royalty rate is more than half again what a bookstore pays the distributor for a print book.

The benefits of success are real.

Unfortunately, I think they're cherry-picking their data.

The top 7,000 best-sellers even in genre fiction make up less than 1% of the total number of books available on the Kindle store. The top 1% are always going to be successful, if they weren't I'd strongly advise staying far far away from the industry.

If you look at Howey's estimated sales figures, he shows that approximately twice as many self-published authors make at least $10,000 a year (his lowest bracket) as big five commercially published authors. The obvious inference, and the one he wants us to draw, is that self-published authors are close to twice as likely to make that amount as those published by the big five.

The problem is that the numbers don't say that.

The likelihood of success is dependent on the size of the field, and we aren't given any indication of those numbers. If the number of self-published authors is smaller than the number of big five authors, then their numbers would look even better; if it's larger, the reverse is true.

All of these numbers need context to be useful, and it's context that we're not getting. We need to see what proportion of the total pool of self-published authors make it into the top 7,000 genre best-sellers and compare that to the proportion of big five commercially published authors that do the same.

That's going to give a better measure for success.

If the numbers are close to even, it's a slam dunk for self-publishing.
If there are ten times as many self-published authors as big five authors, then suddenly twice the absolute number of successes becomes one-fifth the likelihood of success and that's not as pretty a picture.

I want good numbers, I really want good numbers, because I do feel self-publishing is a valid path to success. I just don't think these numbers are as good as people think they are.

[peachiekene]

Originally Posted by Yapyap

There's definitely some element of rating and reviewing differently. I'm not sure how significant it is, or how many readers even pay attention to whether a book is self-pubbed or trad-pubbed, but it's one factor to consider.

I will admit, I do pay attention to whether or not a book is self- or trad-pubbed.
If it's self-published, than the author basically has had to jump through hoops for me to consider buying it. Technically, I rate them higher on average than traditionally published books because
A) I had to do more research into whether or not it was worth buying.
B) I don't acquire as many of them.
C) If it's obvious they didn't get an outside editor, I put the book down and don't bother reading past the first chapter. (Thus no review.)

----
Also, I do know that series tend to average higher in later books. I thought it was implied that most people wouldn't read an entire series if they hated the first book. ^^;
-------

As for Goodreads ratings system. It's not perfect, but I've found it to be more accurate at helping me choose books I will like compared to sites that are actually selling the books. Which is why I mentioned it.
I do think whether or not a site is making a profit from good reviews DOES have an effect on how the book is rated.

As for allowing ratings before a book is released.
I do believe goodreads official stance is or was "rating their anticipation for the book". Which is ridiculous.
But, as others have pointed out, if you get rid of it completely - you're stopping people who are reviewing eARCs from sharing their opinion.
-Although, if you look, it's really easy to tell who is actually reviewing an advance copy and who is just excited about the book.