Strategy Resources Structure Style Layout Composition Production Infrastructure


E-Book Formats Explained Image Formats Explained Advanced Styling & Layout Better Photography Working with SVG Unicode and All That Troubleshooting


Revenue Calculator Contact
Copyright © Dodeca Technologies Ltd. 2016


There are a number of ways that you can produce an e-book, some preferable to others, and there are a number of pitfalls that await the unwary and which can complicate matters. Moreover, there are certain criteria that you must satisfy if you are to avoid a guarantee of failure in your publishing endeavours.

Accordingly, this chapter explores those stipulations, giving close consideration to the issues of pricing and marketing, before considering the favourable and unfavourable approaches to the production phase itself. The next chapter considers the resources that the production process requires, while subsequent chapters examine the techniques you must bring to bear when navigating that process.

Six Rules

The six essential criteria that you must satisfy in e-book production and publishing are considered below. Note that satisfying these requirements does not ensure success, but you do give failure the upper hand if you let one or more of them slip:


Address Demand. Many people with publishing aspirations make a fundamental mistake in writing the book that they would like to write, rather than writing the book that a significant proportion of the world would like to read. Without attention to this, the two factors will coincide only if you are lucky, and the fact is that the success of a book in the marketplace depends upon getting people you do not know to hand over their cash in return for your product. If the market for your book is diminutive, satisfying all other criteria will make little difference to its commercial success.


Produce Professionally. Amazon states that complaints concerning typographical errors, such as spelling and grammatical mistakes, are the most numerous they receive, and the customers in question have a point. Why should they spend money on your book, and bother to read the thing if you cannot be bothered to do a professional job? A poorly produced book is the sign of a rank amateur, and invites poor reviews inevitably, which harm sales.


A Marketable Title. Along with the cover, the title of your book matters greatly, as those two elements are the sole points of first contact with your customer; as with the bait on an angler's hook, they are your sole chance to snare that individual's attention. Good titles – their wording and presentation – possess a number of important attributes, which Chapter Six, Composition, explores in depth.


A Marketable Cover. As unfair as it may seem, people really do judge books by their covers, metaphorically and literally. Moreover, as pointed out in Rule Three above, the cover is the only other point of initial contact you have with your customer. It is the only other mechanism by which you attract customers' attention initially; the only other chance you have to make them take a closer look. It follows that giving your book the most attractive, engaging and professional-looking cover possible is essential.

The fact is also that, aside from actually writing the book in the first place, the creation of non-trivial cover art presents the majority of e-book producers with the most complex and time consuming task in the production process (because most books are of a non-technical nature, and so do not involve charts, tables and images).

However, the positive attributes of such artwork need not be expensive or difficult to attain, and if your cover art will employ one or more photographic components, you can consult Appendix D, Better Photography,​which gives guidance on a clutch of essential optical and photographic principles. Moreover, chapter Six, Composition, covers,​among other things, the design decisions you make when creating your cover art.


Price Competitively. It is difficult to estimate in advance, and with any great confidence, the maximum sum that most people will pay for a given book, but the biggest mistake that you can make is to set a price that reflects your view of the product's worth, rather than the market's view. You may think that your book is worth, say, £10, and this may indeed reflect accurately the hard work that you have put into it, but customers do not care about that; they want a good read for a reasonable price.

Moreover, other small, independent publishers understand this point well, and so compete heavily on price, meaning that you have to be equally competitive if your book is one among a number of similar titles. However, the question of pricing is complex, as it involves a number of factors and the next section below considers these comprehensively.


Do Some Marketing. Few people possess a true talent for selling, and of those who do, some develop entire careers around the pursuit, relishing, it seems, the fact of the hunt over the nature of the hunted. However, for people whose creativity lies along different lines, the idea of mounting concerted marketing campaigns can be discomfiting. Nevertheless, as the anecdotal evidence given in a section below shows, you must sell your book as well as possible in the page that a given retailer provides for it on their web sites (known as the ‘product page’), irrespective of any other marketing you do.

With these criteria established, let us now explore the issues of pricing and marketing in more depth before considering the optimal approach to the production process.


Clearly, the technological aspect to e-books places downward pressure on prices. People spend a significant sum on their e-readers, and so hope to recoup that cost by buying books inexpensively. People are also aware of the essential economics of e-book publishing – there is no physical print-process, and delivery of the product to the consumer is trivial. Given this, and with considerable justification, they expect e-book pricing to be relatively inexpensive in comparison to their print-based cousins.

When arriving at a list price for your book, one obvious approach is to examine the prices for which other books in the same genre sell. You must then to make a judgement as to where to position your product amongst the competition.

You can sell at a relatively higher price, in the hope of generating reasonable revenue at the risk of selling fewer copies. This position has some merit, as long as you do not price yourself out of the market completely, but it limits the number of people who will buy, and thus reduces the number of reviews (positive or otherwise) you are likely to gain in a given period. Personal recommendation (which is what a good review amounts to) goes a very long way to securing sales in any walk of business – a good reputation must be one of a product's most potent assets – and so when people post positive reviews of your book they are doing a lot of valuable marketing for you.

This means that it is favourable to price more competitively, and thus invite a greater number of sales, which in turn will increase the rate at which people post reviews. Assuming the reviews are generally good, this will increase the number of copies you sell, thus recouping some or all of the revenue you lost in pricing lower.

Naturally, however, you should avoid pitching your book at too low a price. A very low price may increase sales dramatically, but too low would reduce rather than increase your net profit. There is also the argument that publishers should have and show some confidence in their publications. If you pitch a book at just £0.99, then you risk communicating an impoverished view of your product (unless, say, the book is relatively small).

These points aside, there are further issues that you should consider, such as VAT (the technicalities of which are covered​in Chapter Eight, Infrastructure). Two broad points apply here:


Prices are VAT Inclusive. When you set your list price for a book, you are setting the figure that you wish the buyer to pay, although this may not be the actual price paid in a given instance, as a retailer may change that in order to maximise its revenue. Whatever the raw value of a given sale, however, it includes the VAT part. This means that the revenue the retailer receives for that sale is the sale price minus the VAT component.

For example, if Amazon sells a copy of your book for £2.00 in the UK, where the VAT rate is 20%, the actual revenue it receives is calculated as follows (where ‘R’ symbolises revenue and ‘VAT’ is a proportion between zero and one):

R = 2.00 / (1 + VAT))
  = 2.00 /  1.2
  = 1.67    

In other words, Amazon will not take the price you set, and then sell the book for that plus the VAT percentage.


Buyer's Location. The VAT that is due on a given sale of a given e-book depends on the buyer's country of residence (it used to depend on the country in which the vendor was based). In Germany, for example, the VAT due on e-book sales is 19%, whereas it is a sizable 27% in Hungary (the highest VAT rate on e-books in the entire EU), which means that you lose a greater proportion of your revenue for a book that sells in Hungary than you do for the a book that sells for an equivalent price in Germany.

This means in turn that you should consider setting the prices for each country so that they compensate for the variation in net revenue to you.

VAT aside, Amazon gives you a choice over whether to sell under its 70% or 35% revenue model (called ‘royalties’ incorrectly, because it implies that Amazon is the publisher – that company is simply a retailer like any other). Under the 35% model, there are broad pricing limits that depend upon the country in which a sale occurs, and upon the size of the file that constitutes the book itself (the ‘deliverable‘). The table below gives the fine details of the pricing bands as they stand currently, and note that, as a very rough guide, a novel of around 70,000 words weighs-in at around half a megabyte (or ‘0.5MB’), although this depends heavily on the cover-image file size:

Up to 3.00MB 3.MB to 10MB
Min Max Min Max Min Max $ 0.99 200.00 1.99 200.00 2.99 200.00 R$ 1.99 400.00 3.99 400.00 5.99 400.00 $ 11.99 2,500.00 23.99 2,500.00 34.99 2,500.00 $ 0.99 220.00 1.99 220.00 3.99 220.00 ¥ 99.00 20,000.00 99.00 20,000.00 20,000.00 20,000.00 £ 0.99 150.00 1.49 150.00 1.49 150.00 $ 0.99 200.00 1.99 200.00 2.99 200.00 0.99 215.00 1.99 215.00 2.99 215.00 0.99 215.00 1.99 215.00 2.99 215.00 0.99 215.00 1.99 215.00 2.99 215.00 49.00 10,999.00 69.00 10,999.00 99.00 10,999.00 0.99 215.00 1.99 215.00 2.99 215.00
0.99 215.00 1.99 215.00 2.99

With this point established, you can calculate the revenue you receive under the 35% scheme using the following formula:

R = 0.35 × (SP / (1 + VAT))

…where SP is the sale price (again, not necessarily the list price you set). Given this, for a book that sells at £3.50 in the UK, where the VAT rate is 20%, the calculation is as follows:

R = 0.35 × (3.5 / (1 + 0.2))
  = 0.35 × (3.5 /  1.2)
  = 0.35 ×  2.91
  = 1.02

In view of this, the table that you see in the sidebar shows the net revenue you receive for a given sale price (up to £10.00), for a book sold in the UK.

Turning to the 70% model, there are two principal differences between that and the 35% scheme, which are as follows:


Single Price-Band. There is a single price band, where the lowest possible price for an e-book in, for example, the UK is £1.99 and the highest is £9.99.


Delivery Fee. Amazon deducts from the net revenue it receives per sale the cost of delivering the book to the customer (i.e. the revenue after deduction of the VAT component). This fee is proportional to the size of the deliverable, rounded to the nearest kilobyte (1kB – one thousandth of a megabyte), and it varies depending on the country to which the book is delivered.

In the light of these two points, the table below give the price bands and delivery tariffs for the various territories (where DF stands for Delivery Fee):

DF/MB Lowest
Highest $ 0.15 2.99 9.99 R$ 0.30 5.99 24.99 $ 1.00 34.99 149.99 $ 0.15 3.99 11.99 ¥ 1.00 250.00 1250.00 £ 0.10 1.99 9.99 $ 0.15 2.99 9.99 0.12 2.99 9.99 0.12 2.99 9.99 0.12 2.99 9.99 7.00 99.00 399.00 0.12 2.99 9.99
0.12 2.99

Given this, the formula for calculating the net revenue that you receive per sale under the 70% model is as follows:

R = 0.7 × ((SP / (1 + VAT)) - DF)

For example, a book that is 2MB in size, and which sells at £3.50 in the UK (the same as the example given above for the 35% scheme), the calculation is as follows:

R = 0.70 × ((SP  / (1 + VAT)) - DF)
  = 0.70 × ((3.5 / (1 + 0.2)) - 0.2)
  = 0.70 × ((3.5 /  1.2)      - 0.2)
  = 0.70 ×  (2.92             - 0.2)
  = 0.70 ×   2.72
  = 1.90

Clearly, this means that you will earn more in this example under the 70% scheme than you will under the 35% model, and this might prompt you to think that the 70% option is always preferable. This, however, is not necessarily the case, as the delivery-fee component means that bigger books generate a proportionally smaller profit for a given sale price. The table below shows this trend by giving the net revenue for a given sale-price against a given book-size, for a book sold in the UK:

£1.99 1.09 1.02 0.95 0.88 0.81 0.74 0.67 0.60 0.53 0.46
£2.50 1.39 1.32 1.25 1.18 1.11 1.04 0.97 0.90 0.83 0.76
£3.00 1.68 1.61 1.54 1.47 1.40 1.33 1.26 1.19 1.12 1.05
£3.50 1.97 1.90 1.83 1.76 1.69 1.62 1.55 1.48 1.41 1.34
£4.00 2.26 2.19 2.12 2.05 1.98 1.91 1.84 1.77 1.70 1.63
£4.50 2.56 2.49 2.42 2.35 2.28 2.21 2.14 2.07 2.00 1.93
£5.00 2.85 2.78 2.71 2.64 2.57 2.50 2.43 2.36 2.29 2.22
£5.50 3.14 3.07 3.00 2.93 2.86 2.79 2.72 2.65 2.58 2.51
£6.00 3.43 3.36 3.29 3.22 3.15 3.08 3.01 2.94 2.87 2.80
£6.50 3.72 3.65 3.58 3.51 3.44 3.37 3.30 3.23 3.16 3.09
£7.00 4.01 3.94 3.87 3.80 3.73 3.66 3.59 3.52 3.45 3.38
£7.50 4.31 4.24 4.17 4.10 4.03 3.96 3.89 3.82 3.75 3.68
£8.00 4.60 4.53 4.46 4.39 4.32 4.25 4.18 4.11 4.04 3.97
£8.50 4.89 4.82 4.75 4.68 4.61 4.54 4.47 4.40 4.33 4.26
£9.00 5.18 5.11 5.04 4.97 4.90 4.83 4.76 4.69 4.62 4.55
£9.50 5.47 5.40 5.33 5.26 5.19 5.12 5.05 4.98 4.91 4.84
5.76 5.69 5.62 5.55 5.48 5.41 5.34 5.27 5.20

As you can see, the net revenue to you for a book that sells for £1.99 is a miserable 46 pence for a 10MB deliverable. At that price, it drops to zero for a book that is 16.6MB in size, and for a book that sells for £9.99 (the maximum price possible) it drops to zero at 83.2MB (admittedly, a very large deliverable).

However, revenue remains constant irrespective of file size under the 35% model, and so there is a point along the size axis where a book of a given price yields the same revenue under both models. That point, for a book that sells in the UK for £1.99, is 8.2MB, and is 41.6MB for a book that sells for £9.99.

Beyond those points, the 35% option delivers ever more, and the chart below shows the difference in the revenue that the two schemes generate. That is, the height of each block in the chart indicates how much more the 70% option gives you over the 35% option for a given price and file size. The area in which the difference goes negative (the flat-topped, triangular zone to the bottom-right) is the space within the distribution where the 35% option has the game.

Three dimensional plot of the difference between Amazon's seventy percent and thirty-five percent revenue model across book file-size and book price

These points are significant because the KF8 format allows for content such as video and audio, and the files comprising such content can be many megabytes in size. It follows that, for those e-books that contain substantial amounts of such content, the 35% revenue option will be the sole profitable option. Given this, it is to your advantage to know the exact size of the file that customers receive when they purchase your book, as you can then factor that into your pricing policy. Accordingly, Chapter Seven, Production explores​the issue of determining file size.

Note here that this guide provides a revenue calculator where you may enter arbitrary values for price, size and VAT percentage, and where all relevant values that result from those inputs are calculated automatically for you. It also works in reverse, in that you can enter a value for the net revenue you are seeking (for a book of a given size etc.), whereupon the calculator will tell you what the sale price should be.


Given an interest in e-book production and publishing, you may have noticed the abundance of books that purport to give you the optimum marketing strategy. They often have titles along generic lines such as ‘X Ways to Get your E-Book to the Top’, where X is usually some inflated and seemingly arbitrary number. The authors of such books appear to be able to sell anything, even if they have very little to actually sell, and it seems that much of their advice boils down to:

1. If it moves, pelt it with your marketing shtick.
2. Repeat daily.

A little thought, however, shows that, were all publishers to adopt such a carpet-bombing approach to marketing, the technique would lose all value rapidly because potential customers would become deaf and blind to the relentless onslaught.

Far better, it would seem, is an approach that involves creating a quality product (now there's a radical idea), and letting good promotional blurb on the product page, word-of-mouth and good reviews do much of the selling. After all, if you are a serious writer, it is likely that your interest lies principally in getting stuck-in to your next book rather than spending hours each day being very loud, crass and annoying as you trumpet your latest work on innumerable Internet blogs and fora.

Nevertheless, the fact is that you cannot expect to make many sales without some form of marketing, at least initially, as the next two anecdotes demonstrate:

My first experience of e-book production and publishing came when a client of mine asked me to process a manuscript of hers that she wished to publish and sell through Amazon. Initially, I used Mobipocket (a free production tool) to generate the book, which misled me into thinking that the retailer takes the promotional blurb that appears on a book's product page from the Mobi/KF8​file itself (an easy mistake first time around). Duly, I entered the promotional material I had written for my client into that tool, but when the book became available subsequently, we found that the blurb was absent from the product page. I had to leave things like that for the next few days, as I was busy with other things, during which time the book failed to sell a single copy.

Returning to the problem, I realised that you are supposed to enter the blurb into the book-upload page on Amazon, and so I updated forthwith the book's details in my client's account, such that it included the blurb. The first sale came in within 48 hours, and the book went on to be a surprising success.

In this light, the importance of promoting a book properly on its product page was clear, and so we did the same for my client's second book, which also went on to be a success. Accordingly, and when she published her third book, I wrote some marketing blurb that positioned it in the context of her previous works, and which then took three of the cases covered in the book, giving a synopsis of each segment, and providing some positive critical-analysis that compared and contrasted the various themes. The final paragraph contained the mandatory call to action, and that was it; about 500 words in all.

However, my client had become nervous of providing any kind of synopsis on the product page, and wished to use only the first and last paragraphs of the promotional stuff. I proceeded according to her wishes, and we used the following (where names etc. are fictitious):

As a follow-up to her first two successful books, The Accomplice Nexus; and Murder, a Victorian Treatment, Laura Ipsum's Exsanguination is a true gem. If you liked Murder, a Victorian Treatment, you will love Ipsum's third work in which she treats us to another broad-spectrum helping of factual Victorian homicide. Here we are party to in-depth and detailed accounts of murder most foul in the form of thirteen superbly researched and written chapters, with Ipsum's ever clear, readable and articulate style portraying Victorian times vividly. This book is one for all true-crime readers seeking a taste of murders from the past, and is available for just the price of a coffee.

Given the clear popularity of her first two works, Exsanguination cements Ipsum's status as an author of note – one to watch, a writer who is going clearly from strength to strength – and with its excellent depictions and rattling-good prose, this book is yet another for true-crime readers seeking a taste of murders from times gone by.

Plenty of ebullient tub-thumping there, but do you notice the problem? It tells you the author has a new book out; that it is just as good as the previous two; and that the author knows her stuff. Yet it states nothing about the book itself, defining it only in terms of the other books. This is akin to describing a house in terms of the houses that surround it – an entirely accurate representation perhaps, but one that illuminates by dint of silhouette only (notably, an example of ‘negative space’ – see below and see Chapter Six).

Unsurprisingly, in the weeks following its launch, the book failed to sell a single copy, and I became ever more confident that the marketing blurb was the problem. Around the same time, I saw an advertisement on TV that trailed an upcoming film. Watching the trailer, I was able to pick up the general thrust of the story; an idea of the principal characters, and an impression of the bits that promised to raise my pulse. It gave me a very good idea of what to expect were I to go see it, but I realised that I still did not know the actual story. That was impossible to discern, as the trailer alluded to the essential situations and tensions only, giving not a whisper as to how those tensions were resolved.

I approached my client about this, explaining that re-reading the blurb we had used for her first two books left me unable to reconstruct the stories in my imagination, even though I had proofread those very stories from beginning to end, and was the author of both marketing pieces to boot. It was obvious from this that neither of them spoiled the business equation, as potential customers had just a fraction of the information that I had, and so it was silly to think that doing the same thing for the third book would have a deleterious effect. Whatever we did, I asserted, the situation could only improve, and to my surprise, my client responded with ‘Actually, I've been thinking the same thing myself’.

Duly we posted the original marketing piece in full, and a customer bought the first copy within 48 hours. Sales and the number of reviews grew from there, and, to date, the book has enjoyed success to equal that of its siblings.

These examples make it clear that some degree of marketing is essential in publishing, and if that makes you uncomfortable, you have no option but to overcome by an effort of will alone the psychological obstacle that stands in your way. If you desire success in your publishing endeavours, you have no choice but to follow the patterns of other publishers, meaning that you must promote your product. This does not mean you must become an empty cynic, caring not a jot about what you sell as long as it makes money. It does mean that you must apprise your prospective customers of your product's existence, and of what it will do for them.

Beyond creating good marketing blurb, however, it seems favourable to avoid what one might call ‘Crony Reviews’ unless you execute your plan skillfully. It seems obvious: launch your book then get everybody you know (your cronies) to post glowing, five-star reviews on its product page. Gather your legions: your best friend, your spouse or partner, his or her best friend, the person next door, that guy you know in the pub (buy him a drink for the trouble – hell, buy them all a drink). Even your auntie Ada is a candidate; 98 years old, never read a book in her life, but she fancies herself as a silver surfer, so she's worth a shot too.

Ignoring ethical questions, this may work if you have a genuine nugget of literary gold, as genuine customers (i.e. people you have never met) will post their own positive reviews in time, which will disguise the froth that your cronies whipped up initially. However, it will reflect very badly on you if your book does not take-off ultimately, and the following, final anecdote illustrates this:

I was shopping on Amazon for inexpensive novels to read on my Kindle, and came across a modest offering that looked like it was worth trying. It was about four people on board a passenger liner, where all the other passengers and crew disappear overnight. I like stories set on planes, trains and ships, and the premise was interesting; moreover, the work had around fifteen reviews, all of which gave five-stars and effusive praise.

So I bought it, started reading, and continued (‘battled’ might be a better term) to the very end, as I became curious as to how such a piece of worthless rubbish could continue with a straight face. The plot was riddled with holes, improbabilities and inconsistencies; the final resolution was disappointing and left obvious loose ends, and the writing style was dreadful slop – it was memorable solely because it was so achingly bad.

Returning to the product page, I read the reviews again, and realised that they had all been posted within the same narrow period, shortly after the book had been published, and that none of them reflected its true nature. Many of them gushed with praise, when it was clear from having consumed the book myself that the putative reviewers had not actually read the thing. These were crony reviews. The author had done a slapdash job of writing it, and had then rallied his associates to the cause; and it was blindingly obvious to this customer at the least.

A final point on what one might call ‘negative marketing’: some authors reply on their product pages to the reviews they receive, negatively so to poor reviews. At least one author of (extant, currently) books on e-book production has exhibited this pattern, posting somewhat acerbic and personal retorts when people post one- and two-star reviews. Crucially, when a reviewer pointed out an important factual inaccuracy in his treatment of HTML, the author stated incorrectly that the reviewer was wrong, when he could have fixed the erratum, re-published, and kept his counsel.

Such behaviour casts a poor light on such authors, which cannot be good for their sales. The desire for people to see you as a real, approachable human being is entirely reasonable, as is the wish to connect – after all, why do serious writers publish if not to connect? But even if someone posts a good review, is it not better to keep your distance? Moreover, when someone gives unreasonably negative feedback, or assertions that are just dead wrong, is it not better to let things be, and trust that the majority of customers will use their own good judgement to filter such chaff (as it appears they do)?

Even with a paragon of content and production, there will always be some bleak specimen of humanity out there who just didn't get it. Accepting that and getting on with writing another book must surely be the finer tack; and when a reviewer makes a negative but valid, factual point, favour fixing the problem over fixing the reviewer.


You can produce your e-book in a number of ways, and it is obvious that your interest will lie, in part, in the simplest and quickest method that will generate the fewest problems. Moreover, and assuming that you concur with Rule Two, Produce Professionally, your concern lies also with doing the highest quality job possible.

This section explores with a broad brush the steps that producing an e-book entails (beyond the prior generation of its raw content), and discounts the inefficient and error-prone methods that are likely to result in second-rate results. It shows from there the optimum approach, thus laying a foundation for the next chapter, which considers the resources you will need when tackling the various production tasks. Note that this is a very good point at which to consult the​Terminology and Notation section in the the introduction in order to apprise yourself of the technical terms etc. used beyond this point.

At its heart, the production of an e-book requires the creation of a single file that comprises the work in its entirety, and which is the file that retailers sell and distribute to their customers (referred to herein as 'the deliverable'). That file can be in one of a number of ‘formats’, and you should consult Appendix A,​E-Book Formats Explained, now, should you be unfamiliar with such technicalities, before continuing here. The key fact to appreciate, however, is that an e-book is no more than a simple web-site bound up within a single file (and so an e-reader is simply a form of ‘dedicated’ web browser), where, usually, each chapter comprises one ’page’ within that ‘site’.

You have a choice over how to generate your book file, the first option being the use of an automated conversion facility, such as that offered by Amazon. Here you upload your word processor files, whereupon a piece of software processes and combines them to generate the file that, ultimately, people buy, download and consume on their devices. Amazon provides this service in order to remove as many obstacles as possible for self-publishing authors who possess minimal technical skills. Putatively, it also saves time.

This route may seem eminently conducive, but it carries two serious drawbacks:


Kindle Only. It does not generate an EPUB file, but creates only the two file-types that are suitable for consumption on Kindle-flavour user agents.​This precludes publishing your book for non-Kindle e-readers (convenient for Amazon), unless you choose the second production option (considered below).


Unreliable Results. You may find that the presentation of the content in the resulting book file differs from your aims. You may get away with an acceptable result when producing a simple book that contains little fancy formatting, but you are likely to experience problems with content that is of a more structured nature. In this case, resorting to tortuous workarounds to generate what you want may well take you longer than if you had gone with the second option.

The flow chart you see here in the sidebar gives the steps that comprise that second option, and note here that this flow chart forms the backbone of Chapter Seven, Production. Broadly, the process entails converting your book's content into a set of files that are very similar to web site pages, and this involves inserting special control instructions into the content, which you express in a very simple language called ‘HTML’. The resulting mix of control instructions and your content yields a body of ‘code’ (not a term of which you need be afraid; and do note too that doing ‘coding’ does not mean you are doing programming, as that plays no role in e-book production).

HTML is the same language that web sites use to give semantic structure to pages (as opposed to stylistic structure or ‘layout’), and a very narrow subset only of the language applies to e-book production. This is because things such as data-entry forms have no place in books, and this in turn is why even those with only minimal technical understanding can still produce an e-book competently and relatively swiftly.

Having added HTML information to your content, you add a different set of rules using another language called ‘CSS’, which control the styling and layout of the content (again, the same as the CSS rules that web pages use; again a narrow subset of that language, as with HTML, and for an equivalent reason). You then create your cover, after which you bind that and the files containing the content together into a single file (an EPUB) by using a special software tool. You can then upload that file directly to non-Amazon retailers. When publishing via Amazon, however, you need to take one further step by passing the EPUB through another program, which will convert it into a Mobi/KF8 file. This is a single, completely automated step.

This guide is geared towards the successful navigation of that entire process – navigation of the flow chart above – and it is clear that certain requirements flow naturally from the points presented above.

Principally, you need to acquire the software that will bind your content into an EPUB. The most prominent applications that support this, and which are available freely are Sigil and Calibre. They operate in essentially the same way, although Calibre yields a wider scope to Sigil in that it operates as an e-book library manager and an EPUB editor too (along with many other features).

This guide centres on the use of Sigil, given the simpler nature of building an EPUB using that tool, but the EPUB production process is essentially the same in both applications, and so the instructions given in Chapter Seven, Production apply equivalently to both.

The other issue of note is that, in addition to their ability to combine a set of HTML/CSS files into an EPUB, both Sigil and Calibre support text editing natively. One critical decision remains therefore, which centres on how you generate the HTML/CSS version of your book's content before creating the EPUB. The options are:


Two-Stage Process. Import your content into Sigil/Calibre directly, and then convert to HTML/CSS within that application.


Three-Stage Process. Add the HTML/CSS to your content as a discrete step, using a third-party editor, and then import the resulting files in their completed state into Sigil or Calibre.

To many, Option A would seem ideal, but the following reasons show it to be unfavourable:


No Independent Chapter Files. Sigil and Calibres' modus operandum is to maintain a book's content as a set of objects that are internal to the EPUB files that they create. That is, importing your content into these tools copies it into an EPUB file, thus rendering it inaccessible directly from the file system on your computer (the usual drives and directories stuff).

This means that editing the file within Sigil or Calibre that contains, for example, chapter two of your book (in order to, say, fix a typo), does not change the original version of the file that you have on your machine. Both of those tools will simply update a copy that is internal to the EPUB in question, and this has two significant implications of its own:


Compromised Problem-Diagnosis. Modern web browsers provide sophisticated ‘debugging’ facilities that allow you to diagnose trouble, and to experiment with alternative styling and layout schemas interactively and trivially, as explained in Chapter Seven, Production. However, editing a book's content, styling and layout within a tool such as Sigil or Calibre places it beyond the reach of the very-capable debugger in your browser. This is not desirable.


Harder Web Publication. As Chapter Eight, Infrastructure, points out, you may decide to promote your publications on your own web site (and that chapter goes into the details​of setting one up). After all, Amazon gives away roughly the first ten percent of a book via its Look Inside feature, thus placing that part of your content firmly into the public domain, whether you like it or not. Given this, and given that E-Book chapters are simple web pages, you might as well capitalise on matters by offering that initial ten percent of the work on your own site, thus maximising publicity.

In this case, and if you have maintained your book's corpus of HTML- and CSS-formatted content as an independent entity (Option B above), it is a simple matter to take that first ten percent and upload it to your site. Usually, such things require just a little manual editing to make the HTML file in question fit in with the rest of the site, but this is trivial.

However, choosing Option A and producing your book by going straight from word processor to EPUB editor, will force you to export the code for the chapter in question back out of that editor, in order to give it the existence as a stand-alone file that it could have enjoyed all along.


Inferior Editors. The text editors within Sigil and Calibre, while competent, are underpowered compared to the dedicated stand-alone editors that are available, and which possess features that are invaluable when converting your content. The next chapter gives details​of such feature sets.

These points show that Option A restricts you, and can thus cause trouble. In contrast, Option B's three- rather than two-stage process liberates you, thus ensuring that you retain maximum control over all aspects of production, which, whether you realise it or not at this stage, is your prime concern. It follows compellingly that a decent text editor is one of your prime assets in e-book production (and the next chapter considers your choices here).

If, however, the above arguments fail to convince, ask yourself why a great many people who work with HTML/CSS professionally favour the three-stage approach. It is unlikely that they prefer unnecessary challenges; it cannot be because they like frustration; nor can it be some absurd issue of fashion, the ‘done thing’ or abject purism. It should be self-evident that a professional practitioner would choose the three-stage path because, ultimately, it is the simplest and thus quickest in that it gives the greatest freedom and thus minimises unforeseen obstacles.

Professional e-book producers are human beings just like you, thus they seek the optimal path, just like you. They want the smoothest ride, just like you, and so they favour the three-stage process because that is the easiest by far in the final reckoning. It follows that doing it just like they do is to your advantage.

Something for Nothing

The previous section advocates strongly that you generate the HTML/CSS version of your content independently from EPUB generation tools like Sigil and Calibre. A question remains, however, over how you should go about generating that HTML/CSS, and the approach advocated here strongly is that you should do this by hand, using the features of a suitable text editor to expedite matters. However, it is possible to use a class of resource that will create the HTML/CSS code for you automatically. This may appear to be optimal, but any convenience you perceive here is a mirage, as choosing this path is deeply unfavourable. This section considers these resources and the compelling reasons for avoiding them.

The resources in question comprise graphical layout programs such as Dreamweaver and InDesign (both supplied by Adobe), and the ‘Save as HTML’ feature that tools such as Microsoft's Word offer. These allow you to lay your content out as you wish it to appear in the final deliverable, where the application writes out the corresponding HTML and CSS for you, either as discrete HTML files (in the case of Dreamwaver and Word's Save-As feature), or as an EPUB (in the case of InDesign). This approach differs little in spirit from using Amazon's automated conversion service, and, crucially, the ill-informed would assert that this approach removes all pain from the production task because ‘you do not have to get involved with all that code’.

This is utter tripe. In physics, mass and energy are conserved, which means that you can neither create nor destroy them, only turn one into the other, and thus it is with the complexity of e-book formatting. That is, you can achieve a given compositional effect in more than one way using HTML and CSS, such that, were you to ‘convert’ from one approach to another, the net result would always be identical.

The key point to realise here is that some of those paths that yield a given effect entail a very roundabout approach, and so are more complex, while others achieve the same effect, but the code in question is considerably simpler. Nevertheless, there will always be a lower limit to the complexity metric for a given presentational effect, below which you can simplify the code no further. That is, there will always be at least one very simple way (in relative terms) to secure your goal, but achieving it in any simpler fashion will be impossible.

To put this another way: there is a base level of complexity beyond which you can simplify no further without dropping something from the book, and this would require omission of content or of some aspect of the styling and layout. Paragraph breaks, borders, italicisation, emboldening, colouration; something would have to go.

Given that tools such as Dreamweaver and Word's Save-as-HTML feature do no more ultimately than write out HTML/CSS for you, according solely to the ‘instructions’ you give them, they are subject to the very same constraint, and so can go no further below the lower complexity-limit than you can. This is inescapable. It follows that you can never avoid complexity by using such applications, as doing so merely hands that complexity over to the tool in question, which forces you then to work with it at arm's length – akin to the gynaecologist who re-painted his hallway through the letterbox.

This has the following practical implications:


User Interface Overhead. You must deal still with the same complexity challenge, but you make things harder for yourself, as you now have to work with the application's user interface. You believe that you have simplified things, but you have done no more than inject additional complexity into the equation.


User Interface Learning Curve. In the case of tools like Dreamweaver et al, you have to learn the user interface first. This too simply injects additional complexity into matters, yet learning the minimal amount of HTML/CSS needed to format the average e-book is far simpler.


Unreliable Code. You could find that an effect you achieve looks just fine in the HTML-writer you use, and looks fine in a web browser, but that technicalities​prevent it from materialising correctly on an actual e-reading device or application.

In this case, and having surrendered control to a non-thinking automaton, you would then be powerless to make things go the way you want, unless you resorted to editing the code by hand. Avoiding that, however, was the reason for using the tool in the first place. You would realise at that point that you could have saved yourself a lot of time and trouble simply by going for manual editing all along.


Inferior Code Quality. Given fundamental computational-limits, the code that such tools generate will always be barely similar to the equivalent that even an unskilled human being would create. It is pretty nasty to look​at too. This means:


Bloated Code. The machine-generated code will always be larger than the optimum, human-generated equivalent, which will increase the size of the book slightly, thus eroding your profits when selling under Amazon's 70% revenue model.


Obfuscated Code. If you got into difficulty, re: Point 3 above, and resorted to manual editing to fix the problem, you would find it very difficult to resolve matters because the machine-generated twaddle would be so hard to understand.

Indeed, there is at least one book available currently that advocates saving as HTML from Word, and which (get this) instructs you on how to dive into the resulting mess thereafter to fix the problems you have created for yourself. Far better, obviously, is the path that avoids, in the first place, creating a swamp into which you must then wade in order to dispatch the alligators.


Degraded Support Opportunities. If, given Point 2 above and the previous point, you decided to hand things over to someone with greater experience, that person would also find it hard to understand the machine-generated claptrap, and could tell you ultimately (perhaps with understandable disgust too) that it would have been easier and faster to code things by hand in the first place. However, and to reprise the argument given in Point 3 above, the avoidance of coding things explicitly was the very reason for using point-and-click all along.

In just the same way that you lose nothing when converting between matter and energy in physics, you cannot lose book-formatting complexity by working with it at arm's length. Such an approach only shifts the complexity to a more-distant place, thus making things more complex, which costs you more ultimately in terms of time, frustration and money.

Ultimately, the problem with thinking that automated HTML/CSS generation will make things simpler is its implication, in the light of the arguments above, that you can have something for nothing. The reality is, however, that genuine something-for-nothing comes only rarely in the form of compositional techniques such as ‘negative space’, which Chapter Six, Composition, explores.

If, however, you are still clinging doggedly to the idea that automated code-generation really will deliver your goal without complexity, despite the incontrovertible arguments above, consider this: were that the optimum path, professionals across the world would use it en masse, yet this does not reflect reality. If the idea really is sound, those professionals, with all their aptitude, understanding and decades of hard experience, must nevertheless be abysmally stupid. Those poor, deluded fools must have missed something that you, with your sunny myopia born of near-zero understanding, can see clearly. Does that sound reasonable? Of course it doesn't.

Do not go near tools such as Dreamweaver, InDesign and their equivalents, nor options such as ‘Save as HTML’ in Word, as these can only hinder not help you. Just learn a little bit of HTML from Chapter Three (and a little CSS, if needed, from Four and Five), get yourself a decent text editor (as advised above and discussed​in the next chapter), and follow the instructions in Chapter Seven, Production.

It will be easier because ‘you need not get involved with all that code’? Yeah, right.