Untangling the New Changes to Amazon’s Kindle Unlimited — Some Early Thoughts

For those writers who may not be aware yet, beginning July 1st, 2015 Amazon will be rolling out a significant change to the payout calculation scheme for their subscription service, Kindle Unlimited (KU). I will assume most readers are familiar with the basics of KU, but just in case, I’ll lay out the differences below.

(This will be a detailed post, so if you want the Cliff’s Notes version, you’d better pass this one right on by…)

Up until the current time, payouts for a borrow under KU have been paid to an author once a reader reads at least 10% of the borrowed book. The amount of said payouts were fluid, based on how many books KU members borrowed, and how much money Amazon decided it wanted to pony up for the KU pot that month. As a result, the actual per borrow payout rate wouldn’t be determined until sometime after the end of a month (e.g. June 2015’s exact payout rate won’t be known until sometime in July 2015).

Starting July 1st, the payout scheme will change to the following (money quote from the Amazon announcement linked to above):

Royalty payments under the new program

As with our current approach, we’ll continue to set a KDP Select Global Fund each month. Under the new payment method, the amount an author earns will be determined by their share of total pages read instead of their share of total qualified borrows.

Here are some examples of how it would work if the fund was $10M and 100,000,000 total pages were read in the month:

The author of a 100 page book that was borrowed and read completely 100 times would earn $1,000 ($10 million multiplied by 10,000 pages for this author divided by 100,000,000 total pages).

The author of a 200 page book that was borrowed and read completely 100 times would earn $2,000 ($10 million multiplied by 20,000 pages for this author divided by 100,000,000 total pages).

The author of a 200 page book that was borrowed 100 times but only read halfway through on average would earn $1,000 ($10 million multiplied by 10,000 pages for this author divided by 100,000,000 total pages).

Now the changes, and especially the examples given by Amazon in the notification itself, have caused some initial confusion among authors. (Click here to see the epic — and growing — thread about this subject currently ongoing on KBoards.com).

I’ll hopefully be able to clear up some of that confusion here, but I’m not promising anything:)

First, the examples given need to be taken as just that — hypothetical examples only. A quick perusal of the numbers Amazon used will show that they just picked these numbers out of thin air in order to make the math less confusing in the examples. Believe me when I tell you, there is no way KU will be handing out payouts as large as the examples might lead some to believe:)

[pullquote align=”left” cite=”” link=”” color=”” class=”” size=””]My first question was: how do we get to a rough ballpark idea of how much payouts could be now that the determining variable is how many pages (out of the total number of pages read in KU for the subject month) are read in any given borrowed book?[/pullquote]

Let’s use a hypothetical based on an estimate of May’s KU numbers (there’s going to be a little math here, but please bear with me):

– Amazon has announced that they will be adding $7.8 million to the existing pot of $3 million for the KU pool that’s used to determine payout for borrows. This brings the total pot for May to $10.8 million.

– Assuming Amazon will still want to keep the per borrow payout at or around $1.35/borrow, we’ll use that rate as our other variable to estimate how many borrows likely happened in May in the KU program.

– So, that’s $10.8 million divided by $1.35 (10,800,000/1.35),  which equals 8 million borrows for the month of May.

– With that estimate, then we need to know how many pages a KU “book” has on average. Unfortunately, I don’t think anyone really knows what the average page count is for all the  books enrolled in KU, so I did a few calculations based on different assumptions: averages of 150 pages, 100 pages, and 50 pages.

We’d multiply that page count average by the total number of estimated borrows to get an estimate of the total number of pages that might possibly (more on this a little later) be read in KU under the new page count scheme:

Conservative estimate (150 pages average):

8,000,000 x 150 = 1.2 billion pages. Ouch.

Mid-range estimate (100 pages average):

8,000,000 x 100 = 800 million pages. Still painful.

Aggressive estimate (50 pages average):

8,000,000 x 50 = 400 million pages. Now we’re (kinda) talking.

The last part of the puzzle is what Amazon considers to be a “page” for an e-book under this new scheme. Up until the present time, Amazon seems to have been using a calculation based upon something around 330 words per page for calculating e-book page counts, where print versions are not available. In the case of books with both print and e-book versions, Amazon has defaulted to the print page count for both the print and e-book versions. (Another good reason to get those print versions up, kiddies)

Now, with this new change to KU, Amazon will be instituting a new formula for estimating e-book page count that will discount differences in things like font, spacings, and the like that might cause variances in an e-book’s reported page counts.

The germane verbiage from the announcement is shown below:

Kindle Edition Normalized Page Count (KENPC v1.0)

To determine a book’s page count in a way that works across genres and devices, we’ve developed the Kindle Edition Normalized Page Count (KENPC). We calculate KENPC based on standard settings (e.g. font, line height, line spacing, etc.), and we’ll use KENPC to measure the number of pages customers read in your book, starting with the Start Reading Location (SRL) to the end of your book. Amazon typically sets SRL at chapter 1 so readers can start reading the core content of your book as soon as they open it.

This standardized approach allows us to identify pages in a way that works across genres and devices. Non-text elements within books including images, charts and graphs will count toward a book’s KENPC.

When we make this change on July 1, 2015, you’ll be able to see your book’s KENPC listed on the “Promote and Advertise” page in your Bookshelf, and we’ll report on total pages read on your Sales Dashboard report. Because it’s based on default settings, KENPC may vary from page counts listed on your Amazon detail page, which are derived from other sources.

(Emphasis on last sentence mine).

That last sentence there is going to cause a lot of confusion for authors who don’t pay attention to it. What all this KENPC stuff means is that Amazon looks like it’s going to assign an estimated page count based on several factors (some of which they include as examples) and attempt to create some standardization in e-book page counts when it comes to calculating KU payout only. Maybe the new KENPC page count will still come out to around 330 words per page? Who knows. Guess we have to wait until July 1st to find out.

[pullquote align=”right” cite=”” link=”” color=”” class=”” size=””]This standardization means that writers will need to remember that a given book’s KENPC number will almost always vary — sometimes significantly — from the page counts reported on that book’s Amazon listing page. [/pullquote]

This will be especially true for books that have print editions, or books that are illustration intensive (some non-fic and virtually all children’s books come immediately to mind, though there are doubtless numerous other examples).

So, for the sake of simplicity, let’s assume that the old word count average of 330 words per page will still be in play.

Let’s go back to the numbers we estimated above and calculate the per page payout based on the aggregate page count of all the books read in KU for May:

Conservative (150 pages average):

$10,800,000 / 1,200,000,000 pages  = $.009/page

Mid-range (100 pages average):

$10,800,000 / 800,000,000 pages = $.0135/page

Aggressive (50 pages average):

$10,800,000 / 400,000,000 pages = $.027/page

Using these estimates, for May, under the page count scheme, the payout for a 100 page book read to 100% would be either $0.90, $1.35, or $2.70.

[pullquote align=”left” cite=”” link=”” color=”” class=”” size=””]The 64000 dollar question that’s the key to all of this is: what’s the average page count for all the books read in KU for a given month? We can only estimate. I’d guess that the average book in KU averages around 100 pages (and this is not just because it makes the math easier).[/pullquote]

As you can see from the above numbers, assuming the 100 page average, this could mean, at face value, a very significant payday for novelists. Under the assumed page rate of $.0135/page a 200 page novel will net a $2.70 borrow payout, while a 400 page novel (both cases assuming a 100% page read through) would net a $5.40 payout.

It seems highly unlikely to me that Amazon would want to be paying out $5.40 for a borrow of a 400 page novel in KU, but the examples I’ve given at least show some possible ranges.

Some Thoughts

We don’t yet know what Amazon’s true goals are with regard to these changes, but we can make some educated guesses based on their past behavior. It’s difficult (and probably ill-advised) to draw any concrete conclusions based on this change alone, but it is useful to think about what this change might be intended to address.

Below, I’ve included some thoughts both from the reader perspective and the writer perspective. What follows are my opinions only.

What This Might Mean For Readers

I think this will probably (in the short run) have little visible change on a reader’s experience in KU. It will take time for the impacts of these changes to filter through to existing authors and publishers, so there may be no large changes evident, at first. In the long run, this change should significantly improve both the selection, number, and most importantly, quality of the novels available in KU. Depending upon how much Amazon wishes to sweeten the pot for novelists, it’s reasonable to assume that more novels will be put into the program, at the very least for one 90 day term (perhaps the opening three months, before publishing the novel to other retailers).

[pullquote align=”right” cite=”” link=”” color=”” class=”” size=””]This should eventually wash out a lot of the 10 page “scamlets” (one of the many obvious — and ruthlessly exploited — ways internet marketer types use to game KU) that both readers and authors have long complained about. [/pullquote]

This should improve the browsing and discoverability experience for readers in KU. We all hope, anyway.

What This Might Mean For Writers/Publishers

I’ll start with a disclaimer here. I’ve been on record with stating that the exclusivity clause of KU is a non-starter for me, as it’s a direct (and effective) assault on Amazon’s competition. That’s business, and I don’t fault Amazon for trying — but I don’t want to do something that even indirectly threatens to reduce the number of retailers I might choose to sell through. Exclusivity clauses are good for Amazon, but they are very bad for both writers and readers. I won’t belabor that point any further here, but if you want to read my full take on it, read here.

[pullquote align=”left” cite=”” link=”” color=”” class=”” size=””]For authors, this will make KU immediately more attractive for full-length novels. [/pullquote]

Depending upon the pay rate Amazon ultimately settles with (I’m cynical, so I’m guessing it will settle down to something less than $0.01/page), novelists may suddenly decide to try KU with more of their longer work.

On the surface, this appears to make KU much more difficult for short story/serial writers, in that business/publishing models relying on quick publication of very short works that take advantage of the better borrow rate (compared to a sale at $0.99), may have just been torpedoed. It remains to be seen if this is will spell the death of such a model, but this would appear to be a step in that direction.

I do think writers of shorts and serials that are actually, well, good stories will still do just fine. It’s the ones that are relying on the model that utilizes a pure volume/low quality approach that will definitely be hurt by this new scheme.

What Do I Think?

The more I think about this, the more I suspect this change is almost wholly focused on combating the rampant gaming of KU that’s been going on. The absurd ease with which KU (in its current form) can be ruthlessly gamed is one of the many reasons why I stayed far away (after my initial experiment with it).

I suspect Amazon probably concluded that the model had too many holes to be sustainable in its current form. I do give Amazon a lot of credit for instituting this change, as it seems to be, while still imperfect, the best way to combat the gaming/scams.

That said, I think Amazon might not normally care all that much about gaming per se. However, I think their real concern is, and has always been, the reader experience. I’m guessing that the gaming that was rampant on KU began to seriously affect the readers experience. [pullquote align=”right” cite=”” link=”” color=”” class=”” size=””]Perhaps the readers grew sick of wading through the oceans of 10 page scamlets that would pay out as soon as a hapless reader opened the first page of a borrowed book?[/pullquote]

Perhaps Amazon is alarmed at the comparative dearth of novels currently available in KU?

Perhaps Amazon is dismayed that so many big publishers decided to take a pass on KU and instead threw in their lot with Scribd and Oyster? Both competitors offer far more attractive terms than KU currently does, so maybe Amazon decided they’d take a step toward competing with those services (dropping the exclusivity clause would be the most concrete demonstration of this, but Amazon will never do it).

I think the main aim though in all of this is to focus on quality … and in so doing play whack-a-mole with the scammers/gamers of the current KU ecosystem. The central tenet of this seems to be a refreshing respect for the sensibilities (and bullshit detectors) of readers. Readers know quality when they see it, and they know shit when they see it too. And this new scheme lets them vote with their feet in both penalizing poor quality while simultaneously rewarding good quality.

This focus on quality, and the utilization of the page count metric is actually ingenious; it encourages writers to write compelling books, rather than polishing those first few chapters, and then phoning in the rest of the manuscript.

For instance, it’s reasonable to assume that there are probably ghost farms that crank out legions of hastily written, poor quality short books (often contracted out to overseas subcontractors). These operations would rely on sheer volume (a few borrows here and there, especially when utilizing the scamlet “model,” adds up in aggregate). My guess is, Amazon grew tired of complaints about the terrible quality of these “books.” They listened to their customers — the readers — and decided to do something about it.

[pullquote align=”left” cite=”” link=”” color=”” class=”” size=””]Does this mean the end of the ghost farm model? No, but it will greatly squeeze profit margins of such “business models.” [/pullquote]

Those operations on the margins will fold and try to find some other way to exploit KU. Perhaps the largest ghost farms will simply try to make it up with even more volume? That remains to be seen of course, but as a reader, I certainly hope they don’t.

This focus on quality will make the ghost farm “business model” much less lucrative; readers borrowing these shitty books will simply drop them like a bad habit, and the payout per shitty book will drop precipitously. At some point, unit sales volume will decrease so drastically, that the people using this “business model” will be forced to do one of three things: greatly increase volume to try to make up for much lower unit sales volume, improve the quality of each “book,” or fold up their tent. I hope many of them do the latter, because their business model (often) negatively impacts the experience for the readers — you know, the people that matter:)

[pullquote align=”right” cite=”” link=”” color=”” class=”” size=””]This focus on quality will negatively impact borrow rings (a practice I find deeply unsavory), but it will not kill them off altogether. [/pullquote]

It will make their scam a whole lot more labor intensive, that’s for sure. Again, the less shitty books that attain increased visibility as a result of borrow rings, the better off the readers will be. It will mean more books will rise based upon quality, rather than because they were artificially gamed up onto bestseller lists.

As an author, I don’t usually care how other authors/publishers do what they do; it’s none of my business, generally, and it has no effect on me. However, when “businesses” utilizing unethical gaming start to negatively affect the experience for readers, the people I care about, the keys to this whole crazy industry, then I very much care what these shady businesses are doing.

And Amazon does too.

Conclusions (such as they are)

Will this make me dip my toes into KU? No — the exclusivity clause is still a 100% deal breaker.

Do I think this will entice some (perhaps many) authors/publishers to dip a toe back into KU. I think so.

Do I think these changes are best for the most important people in this equation, the readers? Most definitely, yes.