Self-publishing is usually described as a way of skipping the gatekeepers. It is more accurately described as taking their jobs. A trade publisher does perhaps a dozen distinct things to a manuscript — structural editing, line editing, copyediting, proofreading, design, typesetting, metadata, pricing, sales into retail channels, publicity, and the unglamorous administrative work of ISBNs, territories and rights. Choosing to self-publish does not remove any of those tasks. It transfers them to you, along with the decision about which ones to pay for.
Most books that fail as self-published projects fail because one or two of those jobs were quietly skipped, not because the writing was bad. So this is about the decisions, in roughly the order they matter.
Two warnings before anything else. Platform terms — royalty rates, price bands, exclusivity conditions, programme rules — change frequently and are widely misreported in articles like this one. I am deliberately not quoting numbers; go to the platform's own help pages on the day you are deciding. And I will not be offering launch-result figures, because the ones that circulate are marketing material.
Is the book finished?
This is the only question that is genuinely non-negotiable, and it is the one authors are worst placed to answer about their own work.
Editing is not one service. Developmental editing addresses structure, pacing and whether the book is doing what it is trying to do; it is the most expensive, the most transformative, and the one most often skipped. Line editing works at the level of sentences and paragraphs. Copyediting fixes correctness and consistency. Proofreading catches what survived everything else, on the final typeset pages.
You can survive skipping some of these. You cannot survive skipping all of them, and the one to skip last is proofreading, because typos are the most visible signal of an unedited book and the cheapest to eliminate.
Costs vary enormously by language, market, length and the editor's experience, so any figure I quoted would be wrong for most readers. What is portable is the method: get several quotes, always ask for a sample edit of a few pages, and choose on the basis of whether the sample made the writing better rather than on price. An editor who has worked in your genre is worth more than a generically excellent one.
If you cannot afford professional editing, the honest alternatives are a serious critique group, beta readers chosen for their willingness to be unkind, and time — putting the manuscript away for long enough that you can read it as a stranger. These are weaker substitutes. They are much better than nothing.
The cover is a signalling device
Authors tend to think of the cover as a picture of the book. Readers use it as a classification instrument. In under a second, a cover tells a browser what genre this is, what tone to expect, and whether it was made by someone who knows what they are doing.
Which means the most common cover mistake is originality. A literary-fiction cover on a cosy mystery will lose readers who would have enjoyed the book. Look at the current bestsellers in your specific subcategory, note what they have in common, and work inside that vocabulary.
Two technical constraints that decide more than aesthetics. It has to be legible at thumbnail size, because that is the size at which most purchase decisions happen. And if you are doing print, it has to work as a full wrap with the spine width calculated for your page count and paper, which is a technical job rather than a design one.
Hire someone who designs covers in your genre. This is the expense I would cut last after proofreading.
Metadata is free and underused
Title, subtitle, series name, categories, keywords and the sales description are the levers that determine whether anyone ever sees the book, and they cost nothing but attention.
The description is a piece of persuasive copywriting, not a summary. It has one job: convince someone who read the first two lines to keep reading. Most self-published descriptions open with backstory and world-building; the ones that work open with the situation and the stakes.
Categories and keywords determine which lists and searches you can appear in. The useful research method is unglamorous: find the books most like yours, look at how they are categorised and what language their descriptions use, and be precise rather than broad. Competing in an enormous category is worse than ranking in a specific one.
And all of this is revisable. Unlike a printed trade edition, you can change the description, swap categories and even replace the cover after publication. Treat the initial choices as a first draft.
Wide or exclusive
The one structural decision that is hard to reverse quickly is whether to distribute everywhere or commit exclusively to one retailer's ecosystem in exchange for access to its subscription programme and promotional tools.
I am not going to state the current terms, because they change. The trade-off is what to understand. Exclusivity concentrates your visibility inside one store's machinery, which can be very effective in some genres — particularly high-volume commercial fiction with readers who consume through a subscription service. Going wide gives you the other retailers, library distribution, and the markets where the dominant store is weak, which matters a great deal outside the US and UK.
The decision is genre-dependent and market-dependent, and reasonable people in the same genre choose differently. What is not reasonable is choosing on the basis of a royalty percentage quoted in a blog post. Read the actual programme terms, including the exclusivity definition and the notice period for leaving.
Print, ISBNs, and the bookshop question
Print-on-demand makes a physical book trivially available and does not make it available in bookshops. Those are separate problems, and the second one is mostly unsolved for self-published authors, because bookshops buy on terms — a trade discount and the right to return unsold stock — that print-on-demand economics do not easily support. A book can be orderable by a shop without any shop choosing to stock it.
Royalty structures for print work differently from ebooks, because printing cost is deducted per copy, which means page count and trim size directly affect what you earn. Amazon's KDP sets out how its print royalty is calculated, including the deduction of printing cost; other print distributors publish their own. Read whichever applies to you rather than a summary, and run the arithmetic for your actual page count before setting a price.
On ISBNs: the retailer will usually assign you a free one, and it will list them as the publisher of record for that edition. Buying your own makes you the publisher, which matters if you want to distribute through multiple channels under a consistent imprint. The economics of buying them vary wildly by country — in India, ISBNs are issued free through the national agency's portal, which removes the cost consideration entirely. Check what applies where you are before assuming it is expensive.
Pricing is positioning
Price communicates category before it affects revenue. A novel priced far below its genre's norm signals something about quality whether or not that is fair; a first ebook priced like an academic monograph will not be bought by people browsing.
The workable approach is to look at what comparable books in your subcategory actually charge, sit inside that range, and understand the royalty bands your platform applies so that a small price change does not accidentally drop you into a much worse rate. Pricing is also revisable, which people forget; promotional pricing and permanent changes are both available to you in a way they are not to a traditionally published author.
What a launch can and cannot do
A launch concentrates attention into a short window, which matters mainly because retailer visibility mechanisms reward concentrated sales. That is genuinely useful and it is not magic. A launch cannot create an audience that does not exist, and the most common disappointment in self-publishing is a well-executed launch to no one.
The things that actually feed a launch are built beforehand: a mailing list, a set of early readers willing to post honest reviews, and relationships in whatever community reads your kind of book. Advance review copies are normal practice and the rules around them are not: reviews must be honest and unpaid, reviewers should disclose that they received a copy, and buying reviews violates every major retailer's terms and is detectable.
If you have none of that infrastructure, publishing quietly and building it afterwards is a perfectly respectable plan. Books do not expire.
The realistic economics
Most self-published books sell modestly. I am not going to attach a number to that, because the honest ones vary by an order of magnitude depending on genre and definition, and the specific figures people quote are usually drawn from surveys of self-selected authors.
What is consistent across accounts from people who make a living at this is that the unit of success is a backlist rather than a book. A second title sells the first; a third sells both. Almost nobody's first book is the one that works, and treating it as a test of whether you can do this at all is the fastest route to quitting.
Avoiding the people who are here for your money
Where there are hopeful authors there are companies selling packages. The distinguishing question is simple and reliable: which direction is the money flowing?
A publisher pays you. A freelancer — editor, designer, formatter — is paid by you for a defined piece of work you could take elsewhere. A vanity press charges you for a bundle and calls it publishing. Unsolicited approaches offering marketing packages, guaranteed placements, film interest or award submissions are, with very few exceptions, targeting authors rather than readers.
Check anyone you are about to pay against the writer-protection resources the professional authors' organisations maintain, and be especially wary of anything that arrives unrequested in your inbox praising a book the sender has not read.
Tags
Taresh Sharan
support@sharaninitiatives.com