Barry & Joe Discussing The Guardian Discussing AuthorEarnings
From Joe Konrath and Barry Eisler:
Joe sez: Yesterday, the Guardian reviewed aspects of the latest AuthorEarnings report in an article called, Self-publishing Surging to 31% of Ebook Market, Claims Report.
Barry sez: I’m a big Guardian fan and think it’s great that they’re covering the data AuthorEarnings has been crunching — data that contradicts a lot of misinformation and legacy industry propaganda. I also think it was entirely sensible for them to reach out to Philip Jones of The Bookseller and Nicola Solomon of the Society of Authors for a contrary view.
Joe sez: Maybe they also could have reached out to you for a supporting viewpoint, since you have written for the Guardian. Maybe they didn’t have time, or see the need. But I see this as a biased piece questioning AuthorEarnings data.
Barry sez: Well, we can always argue about “balance” or whatever in these matters. In general, I like to see a contrary view. My concern about Jones’s and Solomon’s views isn’t about whether they were pro or con; it’s more that they weren’t very coherent.
. . . .
Guardian: According to the Bookseller editor Philip Jones there are “large question marks” about Howey’s data.
“This is a very narrow selection of a particular type of market at a particular time,” he said. “Most people who’ve looked at this in any depth say you can’t extrapolate from bestseller rankings on the Kindle store to a picture of the wider market. Howey sees an ice-cube, and shrieks ‘iceberg’.”
Barry sez: Wait a minute… “a very narrow selection of a particular type of market at a particular time”? They’re tracking over 100,000 titles! But hang on, let’s ask Data Guy himself…
Data Guy sez: The Author Earnings reports are cross-sectional studies of Amazon ebook sales.
The raw data for each is obtained by a custom-coded Java web Spider software program that crawls the thousands of Amazon category Best Seller lists page by page and then “reads” the Amazon product pages for each of the 100,000+ listed books, extracting each book’s title, author, publisher, price, overall Amazon Sales Rank, review counts and scores, and DRM info. It works by parsing the raw html of ebook product pages, following links through all the main and then sub categories. Anything on a product page can be pulled into a spreadsheet cell.
The methodology involves an interplay of four factors: 1) The known overall rank of each title on the Kindle store. 2) The estimated daily sales for those ranks. 3) The sales price of the ebook. 4) And the royalty paid to the author.
1) Is known. 2) Has been compiled by dozens of indies, who note their daily sales at varying ranks, and these rates match up and are regularly re-checked. 3) Is known. 4) Is known for self-published titles and well-estimated for all others.
You can tweak (2) and (4) as much as you want and not break the conclusions found in our reports, which is that self-published authors have taken sizeable market share of ebook royalties on the largest bookseller in the world.
Joe sez: I believe the only way to deny the results of the AuthorEarnings data is to say that Amazon sales rank doesn’t correlate to the number of books sold. Which is VERY presumptuous. When Amazon sneezes, a thousand indies check Jeff’s pulse. If Amazon was tampering with rank, someone would be able to show it.
Almost all media sales follow a power curve. Looks like a hockey stick. You can adjust this curve all you like, but since the distribution of indies to traditionally published authors is even and consistent, it doesn’t change the results. Move a legacy book up in sales by its rank, and the neighboring indies go up. Move an indie book down in sales per rank, and neighboring trad books go down.
What Data Guy discovered, in essence — by looking deeper into the data than anyone had before — was three things:
1) It isn’t just indies in the top 100 taking up 25 – 30% of slots; it goes all the way into the top 100,000+ titles on Amazon.
2) It isn’t just 99 cent works selling at high clips; it’s a range of prices.
3) The difference in royalty rates (5.6x) more than makes up for the difference in price and ranking.
What is the opposing argument? I haven’t heard it from Jones or Solomon…
Link to the rest at Joe Konrath
, Amazon will pay Passive Guy a small affiliate commission on any of your purchases. 
You gotta love how the trad guys point to ISBN data for their analysis but scoff at a methodology that captures real world performance in the biggest marketplace.
Data Guy has managed to take all the BS rhetoric out of the equation and let us see a decent approximation of what’s really going on.
This is a mirror of the whole Trad/ indie thing. They rely on corporate data, flaws and all, while DG is doing it on a shoestring and getting better results.
An excellent fisking, as always.
Using ISBNs to measure ebook sales is one of the goofiest things I’ve ever heard of.
Does anyone outside of the BPHs actually use those any more, especially now that Apple has stopped requiring them?
I like the idea of the ISBNs, but Bowker’s bizarre pricing scheme makes them a non-starter for most U.S. indy authors.
I don’t use ISBNs for my ebooks. Even when distributors required them, I just pasted in the free ISBNs from my CreateSpace editions. That’s quite a few sales that go unaccounted. And I’m supposed to take any industry report about ebook sales going flat seriously.
I can get them for free as a Canadian. I have the account set up and my own prefix but I don’t use them. Given their price in the US I think their relevance will continue to decline as the indie share of the market continues to grow.
Also – getting very tired of this swipe-to-type program replacing random words with ‘Turow’. Anybody else having this problem?
If I could get them for free I’d probably use them. Anything that might help someone find the book is good, right?
Not at Bowker’s prices, though… I bought a block of ten back when, and I think I still have seven left, having only used them for vendors that both required one and didn’t provide one for free (very few places then, almost none now).
Never again.
I chuckled at “Turow”, btw…put me in mind of saying “Beetlejuice”.
Good Lord, Tony, that’s twice. Nobody say his name again or he’ll show up…
Who, Turow?
Aaaaaaagh! Everybody hide!
Bwahahaha
Can you explain this strategy a bit more?
The strategy is that ISBNs have no use for me, and they cost money, so I don’t use them.
Not really a strategy so much as an avoidance of a completely unnecessary step in the publishing pipeline.
Since when does Apple not require ISBN’s? Does this mean that we don’t need to supply them to Smashwords for the Premium Catalogue anymore?
Apple stopped requiring ISBNs a while ago. I can’t tell you when but I can confirm that it happened.
Smashwords is still requiring them, but if you upload directly to Apple you don’t need one.
(And I really wish Mark Coker spent as much time upgrading Smashwords as he did trashing Amazon.)
What if sales rank does NOT correlate with sales? I have my second indie novel which has sold zero copies to date, at a higher rank than my trade (small press) title that has sold.
I don’t know how to sort these numbers. Or how to understand them. Anyone have any insight, or is this a hidden Zon algorithm at work?
How do you get a sales rank without selling any copies?
Deb, I suspect your second indie novel has, unbeknownst to you, sold at least one copy. As Edward rightly points out, books that have sold zero copies are not assigned a sales rank.
Many high-profile indie authors with Amazon sales rankings ranging from the bottom of the charts to the top have collected historic data on their daily sales and corresponding rankings. When they compared their data, all found their numbers to be very consistent with each other, which would not have been the case had sales rank not been based on actual daily sales. Since then, the accuracy and consistency of the daily-sales-to-rank relationship has been validated again and again.
For a couple examples, see:
http://www.theresaragan.com/p/sale-ranking-chart.html
http://kdpcalculator.com
For AuthorEarnings.com, we used these numbers to compute sales-to-ranking conversions.
Thanks for the info, guys.
Trouble is, my dashboard says I’ve sold zero to date on title #2, and yet it has a (very sad) ranking. Could the ranking reflect a foreign sale that isn’t yet reported as such? I do have a small, committed (or maybe should be committed) overseas following.
?