Thursday, 6 December 2018

Withdrawing My Books from Kindle Select

This is just a quick post to announce I will be withdrawing my books from KU/KOLL, in response to Amazon's algorithm changes that mean the KU money is largely now being shared by just two classes of people:
  1. unscrupulous people who game the system
  2. best-selling authors who get millions of pages read per day
My books will still be available for purchase from Amazon, and in print as both A-format and B-format paperbacks. They simply won't be available for free once the Kindle Select renewal periods ends.

For me personally this will drop the royalty payments for the ebooks from 70% to 35%, so I will also be increasing the prices in the next week or so. (I've also been thinking for a long time that I may have been pricing them a little too low.)  This should have a negligible effect on sales, because I am not competing with the scammers to pay for ads, so my books are not being bought.  Very few are even found by people borrowing them (perhaps one or two books per month).

So the ability to borrow my books will end, for each, on the following dates:

Wild Thing - Feb 26, 2019 (The Leeth Dossier #1) (Wild Thing paperback)
Harsh Lessons - Dec 21, 2018 (The Leeth Dossier #2) (Harsh Lessons paperback)
Shadow Hunt - Jan 19 2019 (The Leeth Dossier #3) (Shadow Hunt paperback)

Once my ebooks are no longer exclusive to Kindle Select I'll be able to enrol them in all the other ebook distribution platforms, which do not appear to suffer from the same problems Amazon has engineered for itself.

I doubt I will be returning the books to Kindle Select because I anticipate Amazon will fail to take effective action until the average quality of books available there has fallen noticeably.

For any other authors who wish to withdraw their books, the process is quite simple, as described in the relevant KU help page:
  1. In your KDP Bookshelf page, next to the book you wish to withdraw, click on the "..." (ellipsis) button beside the "Promote and Advertise" button.
  2. In the menu that pops up, click on the "KDP Select Info" item.
  3. On the page that leads you to, clear the tick from the "Automatically renew this book's enrollment in KDP Select for another 90 days" in the panel that pops up. (You might need to click on the Manage KDP Select Enrollment button to get that panel with its check box.)

In happier news, I hope to put the finishing touches to Violent Causes (The Leeth Dossier #4) by mid December (about a week), and launch it a couple of months after that.  I'm pleased with how it has turned out -- with Dave's help (and Jon's) as usual!

Saturday, 3 November 2018

Amazon is failing its authors, its readers, and itself

There’s a growing awareness that using computer algorithms to handle decisions — from debt assessment or even being fired to self-driving cars — can have unexpected and even disastrous effects. Around October 2017, Amazon joined those unfortunate ranks.
Sounds scary? It is, when you dig below the surface a little.
Over a million new books are now published each year. Over half of those are in English.
I started publishing in December 2015, using Amazon exclusively for my eBook editions. Sales and reviews were low but showed a steady growth. My approach was to concentrate on writing each new book and do the bare minimum on the marketing side.
For almost two years that seemed to be working — until October 2017.
At that point, sales plummeted. Had I done something, or said something to upset people? Highly unlikely, since I’d been silently working on the current book. I assumed my silence was the cause. But after discussing sales with a few other authors who used Amazon ads and active marketing, I changed my mind. They told me that from around October 2017, they found if they pay $X to Amazon for ads, they came out ahead by about 10%.
What happened? What follows is my best guess. There are three parts to it: book discovery, Kindle Unlimited, and unscrupulous people.

Book Discovery

Readers can discover books on Amazon through various searches, but in addition there are three important ways Amazon promotes books:
  1. Emails from Amazon to you suggesting books you might like.
  2. Suggestions for other books based on your searches, on Amazon web pages.
Note: included in these recommendations are books that have paid to be promoted.
Best-seller lists. In each category, a list of the books that have earned the most revenue in the last 48 hours.
Popularity lists. Like the best-sellers, but across the last 30 days.

Kindle Unlimited

Readers who pay the monthly subscription fee can borrow books from Kindle Unlimited (KU) at no extra cost. Some of the subscription fees go into a pot of money paid to the authors.
This ‘Cockygate’ article by the Verge notes: “There are over 5 million books available via Kindle, with over a million books available on the Kindle Unlimited system. Amazon is currently paying out approximately 20 million dollars every month to authors on Kindle Unlimited.”
U$20M lying on the table: does that sound like a tempting target for unscrupulous people?
At first, Amazon divided up the pot based on number of books sold. Absurdly short ‘books’ started appearing. Amazon changed the algorithm to count ‘pages read’ instead of ‘books read’. Scammers started stuffing books full of ‘bonus material’ to dramatically boost their number of pages.
It’s been an arms race between scammers trying to grab money unfairly from the pot, and Amazon trying to counter them. This article by Derek Haines argues Amazon’s algorithms can’t count pages or words, but he goes on to add:
KU scammers and ebook stuffers are not the real issues here.
[…] Until Amazon can find a better system to track and record KU ebook reading correctly, the scammers will keep finding ways to make money, and honest authors will lose out.
Or be terminated.
But wait, there’s more…!

How recommendations used to work

Before Oct 2017, Amazon’s algorithms recommended books based on some score related to average review rating, the sales rate, and purchase by other people of similar books. (“People who bought TitleX — as you did — also often bought TitleY.”)
I think this system is still used for the 30 days after a book is released, perhaps extended if the book hits a best-seller list or while it is earning high revenue.

How they work, post Oct 2017

The only books being promoted were those from paid ads. At first, I guessed that Amazon had changed its algorithms to only promote advertised books. There are so many authors publishing, Amazon can profit from authors buying ads.
Amazon don’t disclose how their algorithms work, and rightly so. They’re in a war with those who try to game the system to make money. I have a lot of sympathy with Amazon and what they face.
But since October 2017, the KU pool of money has been paying large sums of money (perhaps most of it) to some unscrupulous ‘authors’. I say ‘authors’ because some are not authors: they are business people who pay ghost writers for content.
The result is, only books with paid ads are being promoted.

Unscrupulous People

David Gaughran’s Kindle Unlimited -- A Cheater Magnet and Kayleigh Donaldson’s Book Stuffing, Bribery and Bullying articles were eye-openers for me — I recommend them. The problem is that payments to unscrupulous people has a terrible effect on honest authors, harms readers, and has started damaging Amazon’s brand (and, I assume, the revenue from self-published books).
The effect on the readers is the smallest: the unscrupulous people can’t afford to annoy them too much, since readers determine the KU payouts. But even for readers, it’s making it harder to find and read books they’ll enjoy.
The unscrupulous people spend big, buying Amazon ads to turn a good profit, paying for those ads from the profits. If collectively they spent 18 million dollars for a 19 million dollar payout from KU, they’d be happy — and hey, that even leaves a million dollars to be divided amongst the remaining 500,000 authors! <sarcasm off/>
The greatest harm is being done to the honest self-published and indie authors. Like most skills, writing follows a Bell curve. A minority are really good; a minority are really bad; and there’s a big spread in the middle between those extremes. (The true picture is more complex, because everyone’s taste differs. But that’s not important for this argument.)
But most good and great writers have no interest or skills in marketing, nor in spending time and effort on it. So most authors can’t compete when it comes to marketing.
If you’re an author selling via Amazon and your readers have declined and your income has plummeted, take heart at least in knowing it’s likely not any indicator of how good your book is. Your small voice is being drowned out by the people gaming the Amazon system. The unscrupulous are producing ten or fifty books to your one, and paying for vast numbers of Amazon recommendations of their books, rather than a meagre few algorithmically-generated recommendations that yours are getting.
Writing a book is hard work. A lot of writers manage one. Producing a book proves to an author they can do it. But without a lot of luck, or a big and skilful marketing effort, few readers will discover the book, so the return is low in terms of money and in appreciation. And with Amazon’s ineffectual response to the cheaters, a lot of new authors will try, but won’t find out if they are great or even good. Unless new authors stumble over this article or one like it, they won’t know the failure may well have been Amazon’s, not theirs.
On top of that, the Amazon algorithms only seem to promote books for the first 30 days. This of course encourages writers to produce shorter books and pressures them to cut corners. For the unscrupulous, it actually plays into their hands since producing content quickly is something they can pay or cheat to produce — they just need to keep one step ahead of Amazon’s lacklustre attempts to rein them in.
By inducing many of the good and some of the best new authors to give up, while still ‘feeding the scammers’, over time the overall quality of Amazon’s book marketplace will decline. At first, not massively. Just a little. But the trend will continue, and in the long term will make Amazon the place for second-rate books which are clearly lower quality than those from other publishers.
The Amazon book marketplace will swell with the people who pay for the ads, and the people with good marketing skills. There’ll be some very good books in there, but on average, the rewards will go to the best marketers and to the scammers, not to the best writers.

How Amazon Could Fix This

It’s in Amazon’s interest to fix this, despite what some cynical people argue. Like other observers, I believe the problem lies as much with Amazon as with the scammers. Amazon seem determined to address the problems with technology (algorithms), not people, and I can see the sense of that: the solution must scale to handle millions of books, hundreds of thousands of authors, as well as defend against sets of active and inventive scammers trying to snatch an unfair share of the money available.
Taken together, this also means that until Amazon develop a solution to the problem, the company appears disinterested and unsupportive, damaging their brand even as the quality of their content declines.
So here’s my set of suggestions for Amazon (the first two of these I owe to a friend, Geoff Dash).
Follow the money. The problem of the scammers is proportional to the amount of money they’re taking. So audit the highest-earning authors in order of the revenue they earn. Examine their books to ensure they’re meeting Amazon’s Terms of Service rather than gaming the algorithms. Add legitimate authors to a register of Verified authors (and even mark their books as such). Add scammers to a list of banned authors and remove them from Amazon.
Suspect the absurdly productive. Amazon could reduce the number of authors they need to Verify by examining them in order of productivity: the authors producing the most books per year (or, per month).
How to Verify. It doesn’t have to be done manually, although the above two principles make that feasible. Instead, deep learning algorithms could be trained with examples of scam content and genuine content. Amazon could use the Mechanical Turk to help examine the books selected for audit. It could build a set of trusted Verifiers. It could use some of its own people to examine the books and explain to the algorithm developers what patterns and qualities to look for.
Only Count Content Once. Use its internal plagiarism detecting algorithms to identify content so that when a reader ‘reads’ a book, that content is only counted once. I imagine something like MD5 markers for a book, stored in a reader’s account. That information could even improve the user experience, by offering the reader a chance to skip the previously-read content.
Be more Transparent. It’s fine to have algorithms making decisions for you, and you need to keep their workings secret to make things harder for the scammers. But have a good and fair system for resolving disputes. Amazon needs to do this because its record of hurting good guys with bad algorithms has damaged its reputation in this respect.
Algorithmically select for quality. By this I mean, engineer algorithms that over time will encourage high quality work by offering reasonable revenue to the author. At the least, stop using algorithms that encourage low quality. The current system that only seems to recommend books published in the last 30 days encourages ‘publish fast’ over ‘publish high quality’. Blocking reviews from professional bloggers and reviewers who are gifted with so many books they don’t spend over $50, is another algorithm that does not select for quality.

What Will I Do?

At present, I see no value in being in Kindle Unlimited, and little value in selling your books through Amazon at all. Being exclusive with Amazon is worthless after the first three months. I recommend anyone who has currently chosen exclusivity with Amazon, to terminate that exclusivity. You may need to wait three months for your current exclusivity period to end, but when your sales are zero there’s no difference in revenue between 35% or 70%. Start publishing your books elsewhere, where you won’t be drowned out by ads and scammers.
If enough authors take this step, it will also send a signal to Amazon. Hopefully, such a market-savvy company will recognise and react to that. Certainly, from reports I’ve read, Amazon doesn’t appear to react to any other signals.
I'm quoting from the Verge’s ‘Cockygate’ article, in which Zoe York (romance author) notes:
“All of these people, without exception, all of them, publish exclusively on Amazon,” she said of the book-stuffers. “This is not a problem on iBooks, on Kobo, on Barnes and Noble, or on Google Play, which are the other major ebook retailers.”

Amazon did a wonderful thing in basically creating and nurturing the ebook marketplace, and through that, stimulating the self-published marketplace. It would be sad if it misstepped now and its know-how and support became irrelevant.
Amazon now rewards good marketers (honest ones, but also the scammers), not good authors. That needs to be fixed, or it will likely lead to the collapse of Amazon as a reputable self-publisher.

Friday, 3 August 2018

Writing, Editing, Creating

I wasn't planning to write a blog article, but felt after yesterday I wanted to make some notes while the recent events were fresh in my mind. I'm not sure what this piece will be about, but I guess it's going to touch on creativity (also see The Creative Process and the Unconscious, Where Ideas Come From) and An update on Creativity and the Unconscious), blocks, editing, and book 4 progress.

Yesterday was one of those wonderful writing days that sometimes happen, so this is partly about what led up to that.

I haven't written any articles for a while because I feel I'm way behind schedule on where I wanted to be for this fourth book. I had felt pretty comfortable that I could have it ready about mid-2018, but that turned out to be wildly optimistic. I'll lay out the reasons why, and let you see what you think of them.

I think I started writing the MS in 1991, and completed that 1st draft near the end of 1993. I just realised that might be a fun image to share… here's a photo of the MS.

I used mostly used sheets of paper that had been printed. In fact, on the left you can find a draft of a page from my wife Stella's PhD thesis on the Percy Folio.

In my first draft of the start of Leeth's story, it opened on the scene where Dr Harmon is in the office of the Mother Superior of the orphanage, looking for a suitable subject to adopt.

Anyway, the MS went through many stages — the next one being typing it up on the computer, in the troff/MM markup. I worked on it and polished and edited it, sending it off to little or no response to publishers, and then about ten years later realised the whole second half was full of story — but no plot. So I bit the bullet and cut it in half. Then it seemed a little short, and I still felt it was lacking something. The Godsson-Disten idea came from somewhere to fill the gap, and I developed that and wove it in to the story. At the end, the MS was back to its original, pre-halved size.

I still kept editing and polishing, improving my craft. I fed that whole MS, chapter by chapter, through the Online Writing Workshop for Science Fiction, Fantasy and Horror, and learned so much from the many very helpful reviews I received, and equally, from writing reviews of other writers' submissions there.

Learning your craft by writing a long novel is a hard way to do it — you need to fix stylistic errors through a much larger mass than a short story — but I rarely enjoy reading or writing short fiction. So I kept at it.

The point of this historical digression is to explain the first couple of special problems I faced with book 4. The first was that I hadn't touched (even read) this half of the original MS in about 15 years, so there was a reasonable amount of work just to apply what I'd learned in the meantime. A bigger problem was the lack of plot. Now, the large anti-emotion plot arc definitely helped, and tied nicely into what was already there — but even so, it needed more. There was also the fear that maybe the old MS simply might not be good enough? I had little idea what to expect when I sat down to read it again, but found myself pleasantly surprised: it was kind of nice, I felt. Leeth had grown and changed more, however, thanks to all the additional trials I'd put her through while developing the 1st three books, so stretches had to be rewritten: what was there was more the Leeth you saw straight out of the Institute: no d'Artelle spirit confrontation, nothing with Disten, and, a really huge difference — no Marcie in her life.

Due to the need to write a short piece for my local writers' group, the Marrickville Writers Corner, I thought that one incident briefly passed over in the MS deserved some expansion: the heartbreaking Luiz assassination. And while writing that piece, while Leeth crept through the dark apartment in shock at what had just happened, she heard a creepy scratching coming from a drawer in Luiz's desk. I had no idea what was coming, just that it connected to Luiz's nasty past, but that led to the black dagger. That tiny seed germinated into the other major plot element for book 4.

The title came surprisingly easily: Violent Causes having several meanings, all of which were applicable for this book, so that was one element that I didn't have a big struggle with. But with all the things that needed addressing, and my desire to keep as much as I could of what was there, somehow even getting to the stage of the MS ready to send off to Dave at seemed harder. I'd arranged a slot, and as the deadline approached I found I was struggling to meet it. In the end I sent it off before I'd done my usual 2nd polish, which was stupid. I'd told myself it would make it easier for me to cut things if I hadn't made over-polished them. Instead, it just made more work for poor Dave.

And when I received his detailed critique, he had some big issues with some of the early sections, and lots of the early chapters had a lot of small issues, too. (Oops: note to self: never skip the polishing!) After some very fruitful discussions back and forth, we got a handle on them and a plan for what I should adjust, and I carried on. And after a month of hard work, I was halfway through Dave's comments, and it was the end of July. Sigh.

There was also one major extra piece of plot to work in, involving Marcie, and a key point in that was to be Marcie's inclusion into one incident. The problem was, that incident was kind of tightly scripted, dramatic, and I could see no way to work Marcie in without disrupting a sequence I'd been looking forward to making concrete, for at least fifteen years. In fact, this scene actually blocked my writing for a whole year, as I wanted to work it out so I could then move on to the next piece of the story. This was before I understood the role of the unconscious, and the simple trick of just writing it down, or skipping past the block. Instead, I'd sit down, and try to work out the scene in my head… and eventually fall asleep, and have used up all the time I'd set aside for writing, with visible results of zero.

Yet here I was, in 2018, with that scene looking and working like a dream — I think it would make a great piece in a movie — and now I needed to add Marcie, in a significant way. So I worked my way closer to this with, not exactly a growing fear, since I had confidence in UTT and my unconscious… but some trepidation, let's say.

And then on Monday I was at the brink: the next thing to do was what I had begun thinking of as The Marcie Problem. Tuesday I visited Mum, and saw the Equalizer 2 with her and a friend (hi Jacqui!), and then had afternoon tea and dinner with Mum, and watched an episode of The Avengers (courtesy of my brother Jonathan: B&W, John Steed and Honor Blackman as Dr Cathy King, not the Marvel Avengers), followed by the Hammer Horror film The Hound of the Baskervilles with Peter Cushing and Christopher Lee. (He seemed tall. I looked him up: he was: 1.96m, 6'5"!)

So, no writing on Tuesday. A whole day with ‘no progress'. But I wasn't too worried, since panic and fear are your enemies and I knew I was just giving my unconscious more time to come up with something good. And it's always a great day, visiting Mum; not to mention, relaxing and having a great meal to boot!

So, Wednesday arrives, and still no breakthrough. I could also tell I wasn't going to get it without a bit more positive action. And now we come to the nub of this article, and what drove me to write it. Because this time, I was more sensitive to what was going on in my own mind.

I remembered my own advice, of using pencil and paper to write stuff down so you're not wasting mental energy juggling ideas in your mind while trying to create new ones. So I took my little hand-made A6-sized notepad with its 4-leaf folded ‘pages' clipped to it, and my trusty propelling pencil. There are some nuances there: that style of pencil, using 2B leads, means your letters are crisp and sharp and the physical act is effortless; and the hand-sized pad means it's not flexing and flopping and you're fighting to support it. It just fits in your hand, you can lie back on the sofa, comfy and warm, and think, and write as ideas spring to mind.

All trying to make it as frictionless as possible when the ideas start flowing, you see?

But I still didn't have my idea. What I did have was some issues. So I started writing down stuff, and also thinking. Not rushing myself. Not stressing myself because I wasn't actually writing the story. Just trying to make the solution coalesce out of whatever soup of ideas was floating around in my head.

What started happening was somehow like doing a jigsaw puzzle, except the pieces of the picture didn't exist until I thought of them, and it wasn't two dimensional but three. The depth part is not especially important, but the time dimension was critical. I started with asking "How would Marcie plausibly end up in this scene?"

(By the way, I'm going to avoid talking about this in a way that gives any spoilers.)

Marcie's appearance needed to fit in with a major plot strand we'd agreed on, that involved Harmon. But how would he know? So, that meant Nelson gave him the info. And then Harmon needed to persuade someone else. And that was kind of interesting. And then Marcie herself needed to find out enough to wind up in the scene. So I wrote that little sequence, which worked very plausibly and nicely I felt, and did good things for the plot, tension, and pacing. I now had the background of my jigsaw, in the sense of both the chain of events that led up to it, and of course I already had the tightly-written scene where Marcie would appear.

Her appearance also had to be significant: it'd be pretty weak if she just waved from the sidelines. Dave had suggested Marcie's appearance should cause problems for Leeth, so that was kind of the ‘shape' of one piece. But I still didn't see how to fit her in.

Now, I'd had one thought, too, that disturbed me: "if Marcie is here, and X is here too, then Scary-Y could easily result, and blow everything to pieces." But at the same time, they say you're supposed to place your characters at risk. Yikes! So that was percolating in the back of my mind.

1st Wednesday of the month was my MWC meetup, so I gave a quick polish to a scene I'd written the week before, and headed off for a very convivial and inspiring evening at Where's Nick. But I still didn't have the Marcie Problem solved.

Thursday arrived, and I kept feeling, for some reason, that Marcie wasn't going to go alone into this scene. I kept imagining someone with her. Which just complicated things, I felt, and I didn't want to do it. But in the end, since I was just lying on the sofa and not exactly making massive progress, I shrugged and decided to explore that. I asked myself "How does Marcie go from where she is, to here?" and just started imagining and jotting down some points. That all seemed pretty natural, and that would put Marcie at the required place, with this other person. And their reaction to my smoothly-scripted scene would definitely fit into new-Harmon-plot-item-Z, and advance that.

Then Scary-Y idea popped back in my head.

And I suddenly saw that Scary-Y could easily unfold after the smoothly-scripted scene. And with that, the problem was solved. The solution wasn't to force Marcie and her scene to meld into the existing scene: it was to let that follow on and develop from the scene!

This definitely felt a Eureka moment.

So then I just let the characters do their stuff. All the different parties with their own agendas, colliding and reacting. Let me just have a count ([Spoilers stripped out]): total of 9 parties. And so events flowed, and the drama climbed, and Leeth pushed back, and things started spiralling into dangerous areas. Maybe half of it was just in my head, the other half the occasional explanatory note written down so I wouldn't forget, or a line or two of dialogue. My notes occupied less than two A6 pages of my tiny writing, but I had enough to set off and I was eager to do so. I broke for dinner, but basically wrote from 4pm till 1am, a bit over 4k words. Sitting back, I summarised it in the flush of excitement as "very plausible, didn't change the existing scene much at all, upped the tension, fulfilled more expectations, causes more troubles down the track, and also has some laughs and more character development, all while advancing the plot!"

And it was time for bed. Looking at the page, I saw that I just needed to write the emotional aftermath, and in due course explore the repercussions of what happened. "I'll just write a little bit, to start that," I thought. Then sat back and saw it was 2am. I'd summarise that effort as taking what had happened to a whole new level. Leeth once again surprised me; but this time, herself, too, and even scared herself.

So I ended the day feeling high, extremely pleased with how it all worked out, and of course relieved to have the Marcie Problem solved and the pressure/stress removed.

I won't have a chance to do any editing or writing today, and I'll be taking a day off tomorrow, so no progress then, either. And I still have the other half of the MS critique to work through. But I'm feeling much more relaxed now. Hopefully, there should only be small bumps in the road ahead.

So, what's next? I hope to have my editing and polishing done by around the end of August. Hopefully I'll find a slot in Dave's schedule for a 2nd round critique, and then with any luck a much easier job for both him and myself following that. Mirella is on the job with the cover design (again, with another idea mainly from Dave), and then I need to try to do some proper marketing for a month or two before releasing this one. So my guess is it will be ready near the end of the year.

Better late than half-baked.

Sunday, 27 May 2018

Amazon book tweaks — Categories, Keywords, reviews, and links

Following some discussions with Barbara Strickland (Amorina Rose's blog) about Amazon Categories and Keywords, and remembering some of what Lama Jabr explained in one Amazon publishing seminar, and after reading this article Six Things You Should Be Including in Your EBook (and Probably Aren't) by David Kudler, I decided to revise all three of my books for Amazon.

David Kudler Thanks

What I wanted to do:

  • Make sure I had links to each book in the series so far.
  • Follow the suggestion to add a review link right at the end of the book.
  • Make sure my categories and keywords were good.

It took me about a day to update the first book, four hours for the second, and an hour for the third. Hopefully this article will help you tweak your first book in less time than I took for mine.

The tweaking is in two parts: tweaking the copy of your MS (probably in .docx or .odt format), and editing your ‘book details' on Amazon.


It also occurred to me (largely because Barbara Strickland had recently commented on how attractive the covers were), that a thumbnail of each cover would make a more attractive link than just the plain text title. Like so:

Novels by L. J. Kendall

The Leeth Dossier:

Wild Thing

Harsh Lessons

Shadow Hunt

Violent Causes

(By the third book, I even realised I should include a link to my Amazon author page!)

While doing this, I found Calibre doesn't preserve hyperlinks on images when you convert to .mobi format, which is what I had been using for upload to Amazon. But the developer, Kovid Goyal, said that format was pretty obsolete, and recommended a different one. I tried epub, and that was just fine.

I wanted to put country-independent links to my other books, in each book. Links to the appropriate Amazon store for the reader's country.  I didn't want to do it through Amazon's OneLink system:

merely because I myself am leery of systems that earn money from people clicking on links.  Granted, I think it's only a few cents, and only if they actually buy the product, but still...  Maybe that's just me being weird, but maybe other people have a distaste for that, too.

Image composition on Linux

I spent some time grabbing images for those buttons to make my own.

I then spent 90 mins finding image editing software that would let me combine them together, for Linux. Lots were recommended, but apparently the idea of a canvas on which you simply drop some images and move them around is a bizarre mental model. Well, apart from the Gimp, but I thought if I could find any other program to do that, I'd probably save time compared to working out how to perform even that simple feat using the Gimp's diabolical UI.

In the end, I found the KDE program KolourPaint, which was very simple and intuitive — it was easy to open up images, and copy them onto a single canvas and then move them around.


Before I did any of this, I first contacted Amazon directly to ask if it was okay to use generalised links that pointed to the reader's local Amazon store, instead of to the US one only (say). I specifically asked if it was okay to use The representative responded and said that was fine.

E.g. the country-independent links for my books in Amazon look like this:

Wild Thing:

Harsh Lessons:

Shadow Hunt:

Previously my Afterword asked people to review the book, but didn't offer them any help to do so. In contrast, the Six Things article described how to construct a link directly to a "rate/review" page for your book so the reader just needs one click, with no typing or searching to arrive there.

I also added a note that reviewers will need to login (to Amazon or Goodreads respectively), to be able to write a review.

So the idea is to add a note at the end, saying something like:

"I hope you enjoyed reading this. If you did, consider rating and/or reviewing it. Thank you!"

(You'll need to sign in to your account(s).)

The Six Things article used an image to click on for the reviews, which I think is more appealing and ‘button-like' to click on than just text. So I grabbed some images from Goodreads and Amazon to make the buttons. I feel that's safe, as Amazon owns Goodreads, and I'm only directing readers to Amazon sites. If you're not selling through Amazon, of course you'd need to pick review sites related to your retailer.

Goodreads review

But the link structures for reviews described by the 2013 Six Things article have changed since it was written. E.g. for Goodreads they'd said:

… um, I can't find it, actually. Maybe David didn't spell it out? He does describe how to find the book ID, though. (You can find it out just by clicking on a Goodreads review link and inspecting the URL.)

Anyway, the review link needs to look like:<goodreads-Book ID number>
Just search for your book in Goodreads to determine your book's ID and examine the link. E.g. for Shadow Hunt the book link is, so the book ID is 34823396, and the review link becomes

Amazon review

For Amazon, you determine your Amazon book ID (ASIN) the same way: find your book and examine the URL.

The example book link is pretty correct, except the http has changed to the more secure https. E.g. Shadow Hunt is

But the link the article describes for reviews, "", is now quite different, and the Amazon review link is now coded like this:

E.g. since Shadow Hunt's Amazon ASIN is B06XVH355Q, the URL for its review is:

I also thought it made sense to repeat the links to all the books in the series at the end, since at that point the reader is more likely to know if they'd like to read more of your work.

Amazon Categories and Keywords

The categories & keywords topic is where things get interesting...

Amazon have a good help page about setting them up:

It turns out that everything on that page is essential reading. I had ignored it.  So my keywords were not doing what they should.  E.g. I didn't know you had to also apply one (or two) of your categories as keywords, for certain Amazon searches (category searches?) to work for your books.
Watching the video that's also on that page suddenly made a whole lot of things really, really clear.  E.g. the stuff at the 3min40s mark where she says something like "Words in the title, categories, and description, will already be part of the search criteria: try to say something new".  Also, she says you should play around with the keywords and see which work best — I've never done that!

It's worth watching — maybe several times.

Setting Amazon categories


"Browse categories are the sections of the Amazon site where customers can find your book. Think of browse categories like the sections of a physical bookstore (fiction, history, and so on).

"During title setup, you'll select a BISAC (Book Industry Standards and Communications) code. The codes you choose, along with your selected keywords, are used to place your book into certain categories, or browse paths, on Amazon. After your book has an Amazon listing, your book's category will appear under the Product Details section of your listing. This will be the path a customer can follow to find your book. [...]

"You can choose up to two BISAC codes for your book. We may also assign additional browse categories depending on your BISAC selections and assigned keywords."

The article also includes tips on choosing categories, and you should definitely read it.

Setting Amazon Keywords

Also note the topic "Categories with Keyword Requirements." For sci-fi and fantasy that links to this page:

Now, for each of my books I'd just made up keywords that I felt best reflected it, and entered those. I suspect I'm not following best practice even now, but at least I'm choosing keywords from the above list. I assume these are words which are most used by readers in Amazon searches.

You might be able to discover them by typing a search term and seeing what words auto-complete in the Kindle store?

Here's my before and after list of keywords for each book:

Keywords — Wild Thing, before
Science fiction Urban fantasy
Strong female lead Magic
Near future Abusive relationship
Adult situations

Keywords — Wild Thing, after
cyberpunk Urban
Strong female lead paranormal
horror Coming of age
Adult situations

Keywords — Harsh Lessons, before
strong female lead some adult situations
some humor some violence
assassin covert agency

I also changed the category from Fiction > Science Fiction > Action & Adventure, to Fiction > Science Fiction > Cyberpunk

Keywords — Harsh Lessons, after
cyberpunk urban, fantasy
strong female lead paranormal
spy, spies, assassin, covert, agency bully, bullies, bullying

(I also made the same category change for Shadow Hunt.)

Keywords — Shadow Hunt, before
thriller cyberpunk
mutants wizard
magic coming of age

Keywords — Shadow Hunt, after
thriller cyberpunk
strong female lead magic, wizard, paranormal
spy, spies, assassin, covert, agency coming of age

So, that's it: a bit fiddly, bit not too bad. I don't expect dramatic changes, but hopefully the rate of reviews will climb from nothing (for almost a year) to a handful of new reviews!

To Do

What I've failed to do (yet) is add the attractive-clicky-button links to make social media sharing of your book super easy. I think that might even be the most important thing to do!

The Six Things article describes how to set up sharing links for Facebook, Twitter, and Pinterest.

Tuesday, 27 February 2018

An update on Creativity and the Unconscious

Why have I written this article now? Well, I've been reading and reviewing and doing small fixes to the half of the Leeth Dossier MS I'd set aside in… (goes away and checks some file dates…) yikes: it may have been 2008, 2004, or even 2003!

Anyway, back then, whenever it was, I'd just re-read a print out of the whole MS and been horrified to realise the whole 2nd half had plenty of story and action — but no plot!

So I cut it in half; then realised the 1st half was now a bit short and could also do with some strengthening, and came up with the Godsson/Disten plot. From that point I left the 2nd half lying there, collecting electronic dust.

Anyway, with Books 1 — 3 published, it seemed sensible to go back to the half of the original MS I'd set aside, that followed straight on. So in October 2017 I think, I brushed off the dusty electrons and poured the MS into the word processor and started reading and reviewing. It went very slowly for a while, as I carefully polished as I went, before I realised that was silly: if I did decide I had to throw away some or all of it, such loving attention and effort would only make it harder for me to ‘kill my darlings'.

Excuses, excuses. So I just corrected minor stuff and made notes about issues that would require more effort to deal with. What with suddenly needing to prepare for Christmas — not to mention a years-overdue de-cluttering of my house — as well as allowing myself a holiday and time to spend on some other non-writing projects, progress was slow. I eventually finished my read-through on Feb 21st, 2018.

At the end, I realised my assessment from 10-15 years earlier had been correct. On top of that I found other structural flaws, all of which I felt might be corrected through ‘the right' plot adjustments. I mean, the 1st three books did lead smoothly to #4 (in particular by continuing a major plot arc from them), but overall I felt the plot was weak, with several strands that felt disconnected. It also had pacing problems too, I thought.

Overall though I still liked it, and despite the flaws, felt this could be knocked into a pretty nice shape. In particular, my gut was telling me that the right modifications to the plot would fix most of the issues I saw.

UTT to the rescue. So, with the need for some creative thinking, and having had good success with the Unconscious Thought Theory in the past (see my older blog article: Where Ideas Come From), I turned to it again.

For the UTT, the first step is to steep your rational mind in all the facts. That was what I'd just accomplished by re-reading the old MS.

But the subconscious is hungry for relevant ideas, so as different thoughts about the story and plot occurred to me — during the embarrassingly long period I spent re-reading the MS — I'd made notes I could refer to later.

Write stuff down. That's a key point: don't rely on your memory. If you don't record the ideas, you'll waste mental energy remembering them or worrying about them. Just write them down instead so you can re-read them when you find yourself in a hole.

Clear the decks of clutter. For the intense parts of the creative thinking process — the parts where you make the act of creation your top priority and give yourself time exclusively for that — you need clear mental space for a problem requiring a creative solution.

So because you need to some clear mental space, write down the problem and all the issues you can think of: like constraints, goals, characters involved, what's at stake… etc. etc. — everything you need to take into account. If you think of more problems, simply note them down as you think of them. These are all grist for your unconscious mill.

Don't get daunted. The funny thing is, that although the number of problems might look scary, each issue is just another piece of the puzzle. In a way, it adds to the richness of the solution; it adds to the stuff your unconscious can use to create an arrangement that makes them all fit.

Give yourself space. Get rid of distractions as far as you can: set them aside and give yourself time. Do this as often as you need to, in each attempt to solve the problem or come up with the needed idea(s). In other words: you can solve these problems a little bit at a time.

Try and try again. You don't need to solve the problem in a single session: you can decide, "Hmm, this might help", or "That might be part of the answer" — so, note those ideas down, too. Sometimes the solution will turn out to be a jigsaw that needs to be assembled. Sooner or later though you'll probably start sensing the big picture: and then the pieces of the jigsaw will start appearing in short order.

Trust your gut. Your gut will often tell you "This idea feels right" — trust it. Think about it, and around it. ‘Your gut' is your Unconscious talking to you. Don't ignore it!

Don't doubt yourself. At this stage, where you're not sure how much of the MS should remain for the final book (since it depends on how big a plot change or other correction is), it can be scary. You could easily spend time thinking about worst case scenarios — like tossing the whole MS! — instead of trusting your gut instincts and the hard work you put in.

Focus on what needs to be cut or added to turn your rough diamond into a fascinating piece of jewellery. At this point, negative thoughts are just distractions. So treat them like everything else: boil them down to their essence and add them to the list of Problems To Be Solved by your creative thinking.

Recording ideas and thinking. For me in this process of creation, some of the time I'll do parts of it on the computer. I'll type up notes, grouping related ideas near each other so I'll tend to reread them all when I revisit one; I'll draw up little timelines (maybe with gaps and question marks: "How does the Department discover this?" or "How do the bad guys learn that?"); I'll sometimes make mindmaps so the connections are graphical and not just linear/textual — and I'll include the holes and question marks, and maybe even colour-code stuff.

But some of the thinking I'll do on paper with pencil, lying down on the sofa. Because for some of it, the computer becomes a distraction. But the pencil and paper is important, because as soon as I have even part of an idea, I can write it down to free up my attention from remembering stuff so I can concentrate on creating stuff.

My review/read-through: While I had been (too slowly!) doing my read-through/review, I'd been:

  • Making notes of ideas that might help;
  • For my local writers' group (the Marrickville Writers' Corner — hi everyone!) I'd also written a few pieces that I felt might fit into Book 4 (Violent Causes), or might just be background events not to be included;
  • Making notes of problems;
  • Forming my overall semi-objective assessment of the MS as a whole.

So I used all these things, plus I re-read my quite large file of random ideas and notes, and the fresh ideas I'd had over the last few months. Lo and behold: a couple of months earlier I'd had a key idea which fitted very well with what I needed! It still wasn't quite right, but it felt like it could form the backbone of the plot strengthening, and so it was back to the Unconscious Thought Theory.

There you have it. With all that fresh in my mind, I thought I'd capture some of the process and get off my backside and write this combined UTT update and status report.

Ah… I suppose that means I should probably write a little status report, shouldn't I? Well, the MS now has 90k words, which at this stage I think I'll be keeping. I'm not sure how many extra scenes I need, or how much editing is needed to fix the other issues I've identified, but my feeling is that it might need as little as 10k words of new scenes (maybe 20k?). So I think it won't be too much longer before I can start trying to book an editing time-slot with Dave at! I should know in a week or two.

Monday, 6 November 2017

Sydney Freecon 2017 - L. J. Kendall interview

Discussing writing, self-publishing, and my sci-fi/fantasy series, The Leeth Dossier. With Q&A.

Once again I burdened an innocent member of the audience with the request to video the talk just using my tiny Canon Ixus7 camera, no tripod, just holding it by hand. (Thanks again to Arthur Dimitriou for graciously accepting the challenge!)

Apologies for the background noise from the nearby main road.

Links mentioned:

Luke's Amazon author page.
The Online Writing Workshop for Science Fiction, Fantasy and Horror.
My cover designer, Mirella de Santana.
Luke's facebook page.
Luke's wonderful editor.
Earlier blog articles about the series (e.g.): All About Leeth

Sydney Freecon 2017, Nov 3,4,5.

Thursday, 26 October 2017

Scribus for print edition covers

Here’s the long-delayed article on using Scribus to make the cover for your print edition. It’s not a full tutorial on using Scribus — I’m sure there are Youtube videos covering that, and anyway, its manual and the tutorials that come with it are very good. It’s surprisingly easy to use, amazingly powerful and well-designed, and… I just love it! It’s also free, and works on Mac, Windows, and Linux. And it works so well!

(I wrote it today because someone - hi, Eustacia! - was interested.)

Why new covers, Luke?

Several reasons. I have a small sci-fi/fantasy convention coming up (the Sydney Freecon, Nov 3,4,5), but I have zero printed copies on hand. I also wanted to make some subtle changes to the covers, along with reformatting the internals of the first two. I was also finally ready to make the 4.25"x7" A-format ‘mass market’ style paperback editions. (Simply because I prefer that size for paperbacks.)

A brief digression, before I get to the topic at hand: the internal changes were switching the dialogue quotation style for the first two books to match that of the 3rd. I’d held off for years, hoping the bug in LibreOffice preventing me from doing that would be fixed. In the end I gave up and used WPS to change them all. Of course, pasting the novel back into LibreOffice left the odd block of text in weird point sizes (some side effect of using ‘direct formatting’ instead of styles, I gather), but they weren’t too hard to fix. And having received help from some of the LibreOffice developers to understand page setup, I had created a good and safe couple of templates for 4.25"x7" and 5"x8" before transferring all the text into them.

One other thing, too: dreaming about offering some printed copies to my local library, I realised there was no way for a potential reader to know the story is quite dark, so I came up with the idea of including some quotes from reviews, since a lot of reviewers did comment on that. That meant I also wanted to insert one page in the very front, as you often see in traditionally-published books. For me though it was less for advertising than to warn readers what to expect.

And let’s start with a picture of the sort of errors that can occur if you don’t pay close attention to the trim and guide marks provided by your publisher. Here’s a picture provided by a very helpful person at Ingram Spark (waves to Erica), pointing out several serious placement errors in what I thought was a nicely-crafted cover for the A-format paperback edition of Harsh Lessons. How wrong I was!

What do you need for a cover?

I had the excellent cover files generated for my guesstimated size (page lengths) from Mirella de Santana, including the CMYK-profile Photoshop .psd files. You need a template from your publisher that lays out the exact dimensions for your cover: where the front, back, spine artwork goes, and what are the safe areas to stay inside. (The cover can shift a mm or few in the printing.) So I logged in to Ingram Spark and used their cover creator tool to generate and send me a PDF file for each of my books, and sat down with Scribus to make the new covers.

Here’s what the template for Harsh Lesson A-format looked like:

Now, because I had given Mirella only a guess of the no. of pages, and I was wrong in the end, I had to fiddle things to make them fit the template exactly.

Even if you have the right sized cover artwork, though, Scribus is great to let you prepare the right file for your publisher/printer/distributor. Although some people say you should be able to work in RGB images (the colour space you see on your computer screen, with its red, green and blue pixel elements), because printing presses cost millions of dollars and have a lifetime of decades, we’re still a long way away from being able to send most publishers anything other than the Cyan, Magenta, Yellow and blacK colour space files they need. Which is where PDF/X (which is also called PDF/X-1a or PDF/X-3) comes in. PDF/X is a file that lets you specify what’s called a ‘colour profile’ for the images in your artwork. Ingram Spark require/prefer ‘U.S. Web Coated (SWOP) v2’. You can download the free ICC CMYK colour profiles for a wide range of needs; as with everything Scribus, there are good pages about downloading, installing and using the ICC profiles, such as this one.

Setting the colour space in Scribus

I remember when I first started using Scribus I had trouble enabling the colour management. After some Googling, I found the answer at (Scribus 1.4.2: "PDF/X-3" option still grayed out, how make it work?) Manfred Moser posted with a problem similar to mine. Some time later he made a follow-up post (emphasis mine) —

"Okay, so I figured it out myself. It is important to turn on color management for the current document. This is done in File - Document Setup - Color Management. Changing it in File - Preferences - Color Management like described everywhere I found only affects new documents and not already existing ones."

If you’ve done that properly, when you open a new file in Scribus and go to File->Document Setup->Color Management, you can easily set things up so it will look like this (it’s showing ASUS for the RGB because that’s the display I use on my computer):

Note in particular the ‘Simulate colours on screen’ and the ‘Show colours out of gamut’. If, when you load in an image, it shows large patches of horrible lime green, that means it won’t print properly: unless you click on the image and choose the correct colour profile for it! As soon as you do that, all that nasty green should vanish. And you do that by right-clicking on the image and choosing Properties, clicking on the Image tab in that dialogue box, and choosing the correct ‘Input Profile’ (which in my case is ‘Embedded U.S. Web Coated (SWOP) v2’). That’s all you need to do.

How to use your cover’s template PDF

One of the many great features of Scribus is that it can read and work with PDF files, and understands some of the internal structure, so the PDF ‘image’ is not just one big image blob. So, just open the template file PDF for your cover with Scribus. E.g. in my template file (which I named HL-4x7-284pp-9781925430073-Perfect-template.pdf) here’s some of what that structure looks like in Scribus:

You can select each object and the corresponding picture element will move into view and be highlighted in the main drawing area.

At this stage, I recommend doing a few things that will make life easier:

1) Click about in the Outline dialogue until you find the three parts of the barcode for your book — the white background rectangle, the barcode itself, and the ISBN — and rename those objects to be something sensible. Then copy each item to the top of the page. With the object selected, the Properties dialogue easily lets you move it up and down in the drawing order. (You click on the green Up and Down arrows in the ‘X, Y, Z’ tab.) Don’t change the size of scaling at all: you want the barcode to scan properly!

It’s a good idea to Group these three parts together for convenience. And maybe name the group something sensible — like Barcode!

2) Now click about in the Outline dialogue until you find the parts of the template that show the inner, middle, and outer edges of the front, spine, and back of your cover, and rename them to something sensible.

3) Open the Windows→Layers dialogue and click on the ‘+’ button to create a new layer. That will make it trivial to flick all your artwork to be visible or not, so you can see whether all the important stuff fits entirely within the safe areas. It’s also a great idea to click on the Lock icon for the Background layer to make sure you don’t accidentally move or change anything in your publisher’s template!

Adding your Cover Image(s)

Next, insert the excellent cover artwork for your cover itself. For simple cases you can just create a single ‘frame’ to hold the whole cover image. Or you might need to create a separate frame for the back, the spine, and the front if you messed up and need to move or stretch or clip parts of the image.

You must create an image frame to hold an image: you do that by choosing ‘Insert Image frame’ from the Insert menu. Next, drag a rectangle that roughly sizes the rectangle. You can easily size and place it exactly, by either zooming in insanely far, or by going to the ‘X, Y, Z’ tab in the Properties dialogue and adjusting the Geometry by directly typing the measurement you want (and hitting Enter when you’re ready), or by using the up and down arrows beside each number. There’s also a small ‘Chain/lock’ icon you can click to tie the Width and Height together. Incidentally, the Name field in this tab is where you can type a new name for the object (Using just numbers, letters, spaces, parentheses, dash, or underscore: again, hit Enter when you’re ready).

Anyway, after you’ve placed and sized the image frame exactly right (or before: it’s up to you), you add the image to fill that frame by right clicking in the frame (or on the object) and choosing ‘Get Image...’ and then browsing to the location where your image is. Another insanely great feature of Scribus is that it understands Photoshop .psd files, including the CMYK colour space (unlike the Gimp), so you can directly use your cover designer’s Photoshop file if you thought to ask them to prepare a CMYK version for your print edition.

When editing the image (resizing, scaling, repositioning, etc.), just keep in mind: the ‘X, Y, Z’ tab in the Properties dialogue lets you adjust the image’s frame (the ‘window’ or ‘viewing port’ onto whatever part of the image you choose to make visible), and the ‘Image’ tab in the Properties dialogue lets you adjust the image itself. Again, you can (and should!) use the lock icon on the X-Scale and Y-Scale so you can’t accidentally squash or squeeze the image so it’s out of proportion.

And a super handy feature of Scribus is the ‘Scale To Frame Size’ radio button on the Image tab, which will quickly scale and position the image to good starting value that you can then later changing by choosing the ‘Free Scaling’ radio button instead.

One last thing: check! Use the Outline dialogue to click on those boundary objects for the front and back cover, and the spine. In the main view, you’ll see fine red selection rectangles appear, and you can use those lines to check that all the important parts of your cover are within the safe margins your publisher has recommended.

It was only after receiving the image at the start of this post, today, that I realised this was a really important step which I had failed to do.

Barcode (and QR code?)

You’re almost done! Just use the Outline dialogue to bring the barcode (that you copied and then Grouped together and named sensibly earlier, remember?) to the top of the drawing order so it’s visible, not hidden underneath the image you added for the cover. Make sure it’s still one inch (25.4mm) high. Don’t change its size!

I personally think it’s a good idea for self-published authors to also add a QR code alongside the barcode, containing a useful search string or URL as the encoded text. For this, I use the free software ‘qrencode’. Here’s the command line I used for Harsh Lessons:
qrencode -o QR-code-HarshLessonsBuyGoogleSearch.png -s 20 -d 300 \


which is just a way of encoding a Google search for: "L.J. Kendall" "Harsh Lessons" buy

There are also similar QR code tools with graphical interfaces, like qtqr.

Creating the PDF/X file for the printer

Your printer/publisher, who supplied you with your book’s cover template, will also have told you what kind of PDF file you need to create, but typically it will be PDF/X (PDF/X-1a or PDF/X-3), which means it supports CMYK.

In Scribus, simply select File→ Export → Save as…, and then

Choose the right ICC CMYK colour profile. These things are designed so the printer doesn’t try to put more ink onto the cover than the paper can handle, as well as faithfully reproducing the colours.

So… how hard is it?

I think my first cover, using Scribus, took me a day to work out most of the stuff above. Then, coming to it this time, I wanted to make four covers in one day (Wild Thing and Harsh Lessons, A and B formats).  I discovered I had no Scribus cover for Wild Thing — I had used Inkscape, and not provided a PDF/X!  I remember Ingram Spark at the time warning me that the cover was barely within the limits of the 240% maximum ink coverage, and a few parts slightly over. I realise now that it worked only by pure luck: the colours happened to be that way.

So I started from scratch, after having not used Scribus for several months (and that last time, was just one day in the last year). Even so, creating the cover for the B-format edition of Wild Thing took just 45 mins. For the A-format, I decided that process had been so painless and easy I’d do it from scratch for that one, too. This time, it took 20 minutes! So then I moved on to Harsh Lessons: each of those took me just 15 minutes.  That's right: four book covers in under two hours, since I already had the actual cover designs and the template PDF files.

So, yes: Scribus is powerful, flexible, robust, and once you’ve picked up a few key concepts, really easy to use.

Go forth and Cover!