A month ago, I published an article titled “Lessons learned from a trilogy: an interim report, and Kindle Unlimited observations“. That was the first part of this article. If you haven’t already read it, please click on that link and do so before continuing with this one. Don’t worry. We’ll wait … what, back already? All right. Here goes.
Dorothy challenged me in December 2017 to complete my latest military SF trilogy, “Cochrane’s Company“, in full, before publishing any part of it. I should then publish the component novels in rapid succession, to maintain reader interest.
She noted that authors in other genres had had good results by doing that. Sales of each successive book in their series had both boosted the sales of those that had gone before it, and been boosted by reader interest aroused by earlier books, meaning that sales for the series as a whole were higher overall.
Dorothy was absolutely correct, although not quite in the way she or I had expected.
The individual books were released at four-week intervals: “The Stones of Silence” on May 14th, 2018; “An Airless Storm” on June 11th; and “The Pride of the Damned” on July 9th. This was to keep them inside a 30-day “launch window” from one volume to the next. (I uploaded each volume a couple of days before the launch date, to make sure Amazon got all their ducks in a row in good time, but only announced the launch on the scheduled day. As usual, Amazon did a very professional job. There were a couple of minor glitches here and there, but they were easily sorted out via e-mail. Kudos to Amazon for that.)
Here’s a graph provided by Kindle Direct Publishing, showing sales (top) and Kindle Unlimited reads (bottom) over the almost three months to date since the first book of the trilogy was launched. Open the image in a new browser tab or window, and if necessary maximize it, for maximum readability. (Please note that the last date’s figures are incomplete, as I had to download the image early to allow for formatting, etc.) I’ve deliberately removed the numbers, as they aren’t important for our purposes here. Focus instead on the visual pattern, the rise of sales and borrows over time as each book was released. The release points are immediately obvious to the naked eye.
You can see at once that sales rose a little per volume after each launch, but not spectacularly so. What did rise very strongly were KU “borrows”. The triple “bounce” is obvious to the naked eye, even without numbers. It seems that, once they were aware of the series, KU readers jumped on it, and read each volume in turn (sometimes “binge-reading” all three within a week). That drove the series’ sales ranks higher, and is still doing so, long after I’d have expected the earlier books’ ranks to drop by much more. As I write these words, all three volumes are still ranked in the top three-tenths of one percent of all books in the Kindle Store. Needless to say, I find that very satisfying.
That bears out something I said in the first part of this article:
Kindle Unlimited has now become a dominant force in the Kindle Store, affecting sales performance and rank to a very great extent. It pays less per completed read than a unit sale, but it has many more subscribers who are potential readers, offering the potential to make up in quantity what’s lacking in the “quality” of a full sale price. A large number of successful indie authors on Amazon, including several who discussed the issue with us at LibertyCon a couple of weeks ago, now estimate that KU “borrows” account for half to three-quarters of their income, dominating “straight” sales by a healthy margin. Since a “borrow” equates to a sale for Amazon’s ranking purposes, that makes their books look much healthier in the Kindle Store rankings, attracting more readers due to its/their “success”. Effectively, a book not entered in the KU program may have to outsell a KU-boosted competitor by three to four times in order to achieve a comparable sales rank in the Kindle Store. That’s a daunting challenge.
For example, Dorothy and I spoke with John van Stry at LibertyCon last month. He told us that KU reads now account for two-thirds to three-quarters of his income – and he’s very successful, making a healthy six figures a year out of his self-published books on Amazon. If he (and other authors) are seeing that sort of contribution from KU to the bottom line, it behooves the rest of us to take careful note.
I also noted that the actual sales, in number terms, of each volume of the trilogy weren’t all that different from each other. For example, in the first 30 days after launch, actual sales for each volume of the trilogy were within 10% of all the other volumes. However, KU reads climbed with each successive volume, driving up the overall sales rank (because Amazon counts each KU “borrow” as a sale for ranking purpose). Before long, KU “whole-book” reads (measured as total daily KENP pages read for each volume, divided by the number of pages Amazon estimated each e-book’s print length to be) overtook daily sales for each volume. They continue to increase their lead over sales to this day. That’s dramatically improved the visibility of all three books, and of the series as a whole, for readers searching for new material in their genre. It’s paid off handsomely.
Having covered sales performance, let’s look at other aspects of writing a trilogy like this in rapid succession, and what it does to the work involved.
First off, I found that I had to edit for the series, rather than for individual books, because the plot was a “trilogy plot” rather than a “volume plot”, if you follow me. Therefore, I couldn’t edit Book One and set it aside, then do the same for Book Two, and finally edit Book Three. At every stage, I had to go back and check the previous volume(s) to ensure that their character arcs, plot development, storyline, etc. dovetailed neatly with the later volume(s), and deal with any errors and omissions. This proved to be a cumulative process, and took me a lot longer than it normally does to edit three individual books. I’d say it increased my editing time by at least 50%, and probably almost doubled it by the time I tackled the third and final volume. It was a major pain in the patootie, and I didn’t enjoy it. I’m looking for ways to streamline and simplify the process.
That experience makes it easier to understand an observation by Nick Cole, co-author (with Jason Anspach) of the very successful “Galaxy’s Edge” military SF series.
I’m informed that, in a recent podcast, he said that they spent something like $7,000 per volume on pre-production costs such as cover art and design, editing, layout, formatting, etc. Of course, the duo have reported making in excess of $200,000 last year from their series, which is outstanding performance in anyone’s book (you should pardon the expression). Nevertheless, the only way they were able to put out six novels in the series during a mere seven months last year (!) was to “contract out” a great deal of the “grunt work” of preparing each volume for publication. Their time had to be devoted to creating output, rather than polishing it. That also implies that those six books in 2017 cost them up to $42,000 in “outside” expenses, which had to be deducted from their profits. Fortunately, the latter were more than enough to justify that outlay.
This is an important lesson for all of us wanting to follow the rapid-book-release model. We’re unlikely to have time to do all the preparation that’s needed, over and above writing the next book in the series. I’ve just learned that the hard way. I’ve always done all my own preparation, with the sole exception of cover art and design (for which I’ve paid an average of $250 per volume). That’s about to change. I’m going to have to employ outside editors, and pay them to refine my output while I generate more. If I don’t, I’ll spend so much time editing that I’ll slow down my rate of release. Outside editing will cost more money than I want to pay, but if I can make more money by doing so, it’ll be cost-justifiable. (For the present, I’ll continue to format my own books [particularly since, now that I’ve bought Vellum and found the software so easy to use, I can do that in a matter of hours]; but that may change later, too. We’ll have to see.)
Cole and Anspach’s success is certainly reproducible, if – and only if – we’re prepared to work as hard as they do. Commenting on a Nick Cole podcast, Brian Niemeier noted:
The author as brand is dead. Kindle whales–the voracious readers you need to drive your sales–aren’t primarily loyal to a publisher, or even an author. Instead, they constantly binge on books in their favorite genre. If your next book isn’t there waiting to sate readers’ hunger when they finish your last one, they won’t hesitate to move on to titles by other authors that will scratch their genre itch.
“Scratching the genre itch” is difficult for solo authors such as myself, and even more so for those working a day job in addition to writing. In the past, I’ve aimed for a length of 90,000-100,000 words per novel. I simply can’t produce one of those every 30 to 45 days, even if I write full-time. It’s too much work, not just in the writing, but in the pre-publication processing. I’ll have to drop my target length to 70,000-80,000 words (similar to many of the Cole/Anspach novels, and to John van Stry’s books) if I want a more rapid, sustainable release rate (say, a book every 60 days). I may be able to write a few shorts, to intersperse with the novels. Shorter books will, in turn, mean charging lower prices for them, and other compromises. I also have to see whether my readers will like them, or whether they’re going to hold out for longer novels. If they do, I’m probably not going to be able to maintain a 60-day release cycle. It’s going to be an interesting learning curve, that’s for sure!
That, in turn, may affect where and how we sell our books. There’s a great deal of debate about “going wide” (i.e. selling through multiple vendors across multiple platforms) versus selling on Amazon.com alone (which is also the only way one can get access to the Kindle Unlimited market). The administrative overhead of “going wide” is not inconsiderable. It involves monitoring sales across multiple vendors, planning and running marketing campaigns on each platform, converting currencies, accounting for income and expenditure by platform as well as by book (with all the tax implications involved), etc. The more successful we are, the greater that administrative overhead will become.
As indie authors, running our own small businesses, that’s a significant drain on our time: but, by publishing exclusively with Amazon.com, almost all that drain goes away. One can devote more time to creating rather than managing. As one’s output grows, and “writing time” becomes more important, that bears thinking about very carefully. I note that many successful indie authors have “voted with their feet”, shut down their previously wide distribution, and now publish on Amazon.com alone. I can understand why. Sure, there are potential disadvantages to relying on one vendor. However, if the prospective rewards outweigh those risks… Only time will tell.
There are other considerations that are equally important. Depending on our approach to writing, we may be able to learn from Michael Anderle, author of the Kurtherian Gambit series.
He started with the concept of a “minimum viable product“, and applied it to fantasy/SF fiction, working as fast as possible and devoting minimal time to “polishing” his books. He typically wrote and published them within two weeks from start to finish. He was very successful in terms of sales, progressing from zero earnings to over $10,000 in his third month on Amazon.com. His writing and publishing model is unique in my experience, and it’s not necessarily one I’d like to follow – but it worked, and still works, for him. You can read more here. (Do also listen to the interview with him, and read the linked [.PDF format] summary of his “lessons learned”. They contain useful and thought-provoking information.)
I’m not sure most of us would be comfortable with such an approach – I know I’m not, having read several of Mr. Anderle’s books. However, it’s clear from his success that many readers aren’t worried by it. That means we have to cost-justify our approach, rather than simply assume what we’re doing is the right thing. In fact, taking everything above into account, I’m going to have to cost-justify my decision to write each and every book.
For example, right now I’m busy with the third book in my Western Ames Archives series.
It’s a labor of love for me. I enjoy the Western genre, and I think I do a reasonably good job of writing in it… but it’s a less popular, lower-volume genre, with sales per book that don’t come close to successful science fiction or fantasy. To write a good Western takes a lot of research time, as well as travel to and from potential locations, accommodation there, etc., before I write a single word. Is it cost-justifiable to devote so much time and effort to a novel that I know, even before publication, will not sell as many copies as one in a more popular genre? Will I be forced, by sheer economic considerations, to stop writing in a genre I enjoy? That’s a very good question, and one I’m going to have to answer before much longer. Would writing much more simply, following Michael Anderle’s formula, help to alleviate such concerns, bringing a cheaper product to market much faster? That would mean sacrificing quality and originality. I don’t want to do that… but can I afford not to? (Whether or not the market will continue to be receptive to such products over the longer term also remains to be seen, of course.)
I’m applying all the above considerations to potential future series. I have two in mind right now, one military SF/space opera, one fantasy. Both look set to be at least trilogy-length, perhaps more than that. I also have other, ongoing series to continue or complete. Which is likely to produce the greatest return on my investment of time and resources? Should I choose which to work on based on whether I’ll enjoy it more, or because I’ll earn more? I don’t want to have to be so mercenary in my approach… but I also have to make a living. Cold, clear, level-headed thinking is going to be ever more essential as the market becomes more and more competitive.
I’m even debating whether I should follow the Cole/Anspach model by looking for another indie writer with whom to collaborate on a series. I haven’t done that before, and it may be a steep learning curve for both of us. Even so, if I can find someone whose mind works in similar ways to mine, so that our mental and creative gears “mesh” with each other, it might enable both of us to produce (and earn) more together than either of us could on our own. That’s worth thinking about. (Of course, any collaborator would have to be able to devote as much time to writing as I can, and be willing to work as hard, and contribute to pre-publication costs – up front, before we start work – on a share-and-share-alike basis. Any or all of those factors might be deal-killers.)
Well, friends, those are the core lessons Dorothy and I have learned from the launch of my most recent trilogy. I hope going into this much detail has helped you understand the state of the market at present – or, at least, how we found the state of the market. Based on our discussions with other authors at LibertyCon a few weeks ago, we’re fairly sure that our lessons apply to a great many more people than just ourselves. The question is, how should we apply them? I look forward to your suggestions about that in Comments.
Mad Genius Club 
33 responses to “Lessons learned from a trilogy, Part 2: the impact on sales of rapid releases, and other factors”
Staying fresh does provide value, because it helps you write your other books. Similarly, I have noticed that some writers tend to have a “practice” novel and then a “mastered technique” novel or novels. Obviously it would be better if every book was perfect, but trying new stuff seems to result in longer careers. (Yes, even Barbara Cartland had an experimental side.)
I’m skeptical that collaborating decreases the amount of work that the individual authors have to do.
I’m sure it decreases the amount of writing, but the increased effort of coordination and consistency seem like they’d eat any savings. And then some.
I suspect with two, you could probably save some time, though you wouldn’t come anywhere close to cutting it in half. That is, assuming that they were compatible authors in the first place; results may vary if you try something like a Stephen King/Danielle Steele collaboration.
Thanks for the observations on the difficulties with editing. I’ve been predicting that. I’m just starting a trilogy I plan to rapid release next year. Most of my current books out are around 100K, but I’m aiming for 70k or less with these next ones.
However, having just finished a trilogy, I’ll say that the continuity issues are there regardless of whether one writes them together or a year apart. It’s good I’ve gone back to the earlier books, or there would be some real continuity errors. One just forgets.
Agreed. It is easier when you can write boom, boom, boom, then release more slowly, but the errors creep in even so. (“Oops, wasn’t he a five foot four blond in the last book? Now he’s five ten with black hair and a tattoo? Better fix that.”)
Thanks for this. Very helpful.
On the subject of genre and economics, I recall years ago that Scott Westerfeld wrote the first few books of a space opera for Tor…but he was not able to finish it–because he made so much more money with his YAs that he could not afford to pause and write the space opera.
You may be high enough in the rankings that you’ll maintain visibility between releases even with slower production. Guess you’ll see in the next couple of months, unless you’ve got one ready to go. Short stories are good, for bridging gaps, or even to lead off a series of close releases.
Yes. The marketing books call it being “sticky.”
This isn’t a case of comparing one power tools and manual labor electronics enclosure assembly line with another. The individual variations of mental processing mean that Author A is not guaranteed Author B’s changed results by implementing Author B’s new procedure.
So what I hear is that you need more information about what changes in your process would mean for you. Could you afford to write off the costs from experiments like Sarah’s short story a week exercise? Could you switch tasks between a main project and a faster simpler experiment? How much time do you expect for future production? How much of that is worth sacrificing for little direct profit in exchange for possible greater productivity with the remainder?
Secondly, Pete Grant is a bit of a brand. You’ve ended up targeting a particular market by writing stuff that really works for you. A collaboration might not change that. My untrained gut instinct is that faster simpler production might better fit another market. I very much haven’t been able to pay close attention, but English translations of Japanese Light Novels seem to be successful enough in the Kindle store. As maybe stuff chasing the same market, originally in English. Now, a lot of light novels have elements that you probably would not be interested in. (I recall you saying that you won’t do porn?) The internal art of the typical light novel is unlikely to be in your budget. If you do change your story structure, be sure you understand your new target market, their expectations, and so forth. It is very likely that my specific guesses about such a new target market aren’t perfect.
Thanks for writing this up and posting it – very interesting and informative!
Thank you for the update and congratulations. I’ll admit the idea of pumping out a novel every two weeks and creating the “minimum viable product” seems foreign to me as a noob still struggling with her first book. I’ve gone through two partial rewrites even though I started out with a decent outline. I recently crossed the halfway mark and I’m thinking I might have to go back and chop a bunch because I got carried away with setting and new characters. Its a ya fantasy so readers will be expecting a lot from the setting but I went too far.
Its frustrating because I would like to release a new book in the series every other month, yet here I am over a year since I started in earnest. It feels like that goal is out of reach.
Not really out of reach. You just need to get this first one “under your belt.” Then you will know where you problem areas are, what works really well, and what to keep an eye out for. The next book will come more quickly, and the third more quickly still.
Keep in mind, Peter Grant, Sarah Hoyt, Margaret Ball, Pam Uphoff, me, and many other mad Genii and readers have been at this for a while. You don’t see our mistakes, the books that sold slower than soil creep on a peneplain*.
*A peneplain has reached near equilibrium and doesn’t really erode until something changes the river/stream.
Yeah, this. Basically, with the first stories you are building your factory.
When you ask ‘how long does it take for each car to come off the line’, you don’t measure from when you decide to build a new factory and design a new car to the first car coming off the line. Give yourself some cars to work out some of the early problems. Then, record the time when one car comes off the line, and record the time when the next car comes off the line.
Totally apt.
Ooh nice analogy! I’ve lost track of when I started but I’ll make sure to note an estimate when I’m done. Thanks for the encouragement guys!
It may be out of reach today, but not necessarily forever. As you write more books, you’ll probably have less rewriting to do.
Even in the 80’s, a successful writer (below the bestseller top tier) was generally a writer who’d mastered the art of writing very clean first drafts. It’s a skill. It can be learned.
And finish the story first, then edit. You can cut out parts then, or put them in a new story. But get to the end first.
I’m doing a lot of “Oh this is better than the outline!” and then I have to go back and rewrite part of a chapter to make the thing happen. Today I wrote for 4 hours and ended up with 1,000 words fewer than I started with. D:
Problem with that approach is that it is possible to reach a revision singularity. Where you sink so much time into revising the document document so far to fit the current vision that you never actually finish.
Best practice seems to be to write the thing start to finish regardless of changes in vision along the way. Then revise once at the end. When you have a complete document that wanders through different visions at different stages? Then you can pick a single best vision, and revise everything to fit that one vision.
Imagine a forty chapter story, where the vision changes every five chapters. If you revise whenever the vision changes, you are doing 5 chapters revision, plus 10 chapters revision +… equals the work of revising 140 chapters. Doing it at the end is 35 chapters of revision. Okay, in real life it isn’t so cut and dried.
Other issue, writing is a psychological process of manufacturing. A lot of people have reported that writing a draft and editing are very different processes. And that to maximize productivity, so that the brain knows which it needs to be doing, they need to keep the processes separate. So when they write, they write only, not editing or revising, because those would confuse the brain. The brain says, “you’ve given me the psychological cues that tell me I’m writing, but you are doing a copy edit of the work you’ve just done? What am I supposed to be doing? Am I imagining some bizarro land, or am I checking a bunch of finicky details?” The brain tends to do things best when it is doing a single thing at a time.
“Imagine a forty chapter story, where the vision changes every five chapters. If you revise whenever the vision changes, you are doing 5 chapters revision, plus 10 chapters revision +… equals the work of revising 140 chapters.”
Haha! This is exactly what I’ve been doing! I needed to hear it put that way. I’ve been driving myself crazy since chapter 5. Hopefully I can remember this for next time I get the urge. Thank you.
The math gets much worse as the vision changes more frequently. Every two chapters of 40 is 420. Compared to 38 of just revising once.
With the possible exception of David Drake, our understanding of our story is going to change between the start and the end. Given how almost all of us work, letting ourselves revise before we finish tends result in quitting because we’ve put in so much work that we are sick of it, and document is so confused a mess that we no longer remember why we started.
We get better at writing by writing. I can not tell you how many times I’ve seen someone halt while telling a story, realize they’ve gotten much better, start revising from the beginning, and never manage to finish. It is an easy trap for us writers to get caught in. I’ve had days where I was ashamed of something I wrote in the morning. The advantage you have as an indy is that you are not showing things to the public before you finish, and it’ll be easier to control embarrassment with early weak writing.
Indeed.
Peter, thanks for this post. Your three-in-a-row release schedule generated a sales-build that is obvious and can’t be argued with. I’ve got some work to do. Also the off-loading of editing etc. bears consideration for a volume-production author.
As usual, you’ve given me much to ponder. Thank you.
Lester Dent averaged 6 novels a year for 30 years (though much more during his peak years). He developed his master plot, and used that structure across several genres. The KU market is essentially the same as the early pulp market. There are readers who want exciting stories on a regular basis. There is much financial success to be had by adapting your skills to serve that market. I recall at one time Erle Stanley Gardner dictated his stories, and kept two full time secretaries busy typing them.
Can you write faster by using modern dictation software? Write normally, but use dictation software to create a running chapter synopsis to make editing easier? Can you speed production by writing to standard length, and using a standard cover design within each genre?
The concept of write carefully, distribute quickly seems to be the path to success
This is a lot of interesting things to think about.
On the one issue, I think that it’s probably *economically* wise to write the occasional less profitable book *if* it’s going to keep a person fresh and interested in the profitable ones. The same thing could be said for any use of a person’s time that isn’t “productive”. Hobbies! Or resting on the Sabbath. Etc.
The “fast release” model that everyone is talking about is both super encouraging and a bit discouraging. On the one hand the examples of considerable success always make success seem more possible. On the other, even developing an excellent work ethic won’t give someone with a day job the ability to imitate Cole and Anspach’s production. I wonder if releasing books in series bunches (trilogies or quartets, etc) every year and a half might work better for a slow producer than one book every six or nine months. It’s true that the dry patches would be a lot longer. And this doesn’t seem like something that could be tested either, unfortunately.
And lest waiting a year and a half seems outrageous on the face of it, if a person were submitting to publishers a single book can take twice that long.
Lastly, and very interestingly is the quote by Niemeier that the author as a brand is dead, that people binge read series, not authors. Well, we binge watch television too. Whole series at a time. And I’m led to believe that media tie-in novels have always been bigger than I thought they were, and those provide a similar experience of being in the same world with many of the same people.
Can authors write singles any more?
Quote: On the one issue, I think that it’s probably *economically* wise to write the occasional less profitable book *if* it’s going to keep a person fresh and interested in the profitable ones. The same thing could be said for any use of a person’s time that isn’t “productive”. Hobbies! Or resting on the Sabbath. Etc.
very much this:
I know that David Weber has a similar thing in that his War God’s Own series makes considerably less than his other books, but they provide a nice break for him.
It’s more profitable for you to take a break from Sci-Fi/Military by writing a Western than by not writing at all because you are burnt out.
And it means that all the travel expenses for research are tax deductible 😀
23 skidoo
Heck, I am doing a lot of research for the Mil SF series i am working on. Even tho i scribbled town most of a chapter off of a flash of inspiration, I am finding that filling in the book is taking longer…
checkthebox.
Thanks for the research and update Peter!
I’m just a reader, not a writer, so it’s hard for me to comment about what a writer needs to do to make money. I can comment on what it takes to get money out of me — but the time needed for the writing/editing process isn’t something I know firsthand.
As a reader, the process I go through, when confronted with the existence of a new book, is deciding if it’s worth trying at all. I’ve got hundreds of books in my “To be read” pile, so, unless I think it’s got a reasonable chance of me liking it then I don’t buy it at all. What makes me believe that there is such a reasonable chance comes from (in order of effectiveness): (a) reading enough other work by the author, (b) personal recommendations from people who know my tastes, (c) impressions from hearing the author speak at conventions, (d) the author’s online presence, or (e) general reviews by people whose tastes I think I understand well enough, but which aren’t directed to me personally.
In the case of a multiple-book set, I want to know if it’s really a series, or if it’s a single book which, for commercial reasons, is being marketed as multiple books. For example, Lord of the Rings (as a trilogy) is actually a single book — binding technology at the time wouldn’t let it be printed as a single book. Or, to pick your books, Cochrane’s Company is really a single long book, whereas Maxwell is clearly a series. If it’s a series, I’ll buy them if I have reason to buy the first, and if it’s good enough, keep buying them so long as they’re interesting. If it’s a single book, then I’ll wait until the “book” is finished — so I waited until Pride was out before buying any of them, so I could read the book/story as a single book, instead of stopping in the middle. If it were by someone I was less sure of (past data from reading your other stuff told me to expect I’d like it in this case — but there are cases where I’m a lot less certain), then I’d buy the first one, and treat it like any other book that I start reading — read enough to give it a chance to be interesting, and stop reading when bored — which means if I read well into the first book, and am still interested, I’m likely to buy the rest whenever I have time, so I can read the whole story at once.
I don’t know if that’s useful data for you. I’m likely be closer to the classic SFF reader, in that I read across all subgenres in the field, rather than the more modern picture of readers as being more subgenre-specific in their tastes. So I had no reason not to pick up Bring the Lightning, even though I normally don’t read a lot of Westerns — I knew from your other stuff that you could write well, and many people who can write well in one genre can write lots of things well. And, when I enjoyed it, that meant to keep buying the series as new volumes come out, since you’d showed that, as expected, the fact that you could write well in MilSF reasonably predicted you could write a Western well.
Just curious… when I re-read a book from KU, I know you don’t get any more income from it (assuming I read it all the first time, of course!) Is the addition read counted in your stats?