There is a lovely moment in independent publishing when somebody tells you your book can be available to thousands of bookshops. You briefly see the window displays. The tasteful little recommendations. A bookseller pressing your novel into a stranger’s hands with the urgency normally reserved for an evacuation.

Then you discover that “available” means a shop can order it if somebody asks.

This is useful. It is also several commercial events short of the scene you were enjoying.

Choosing between KDP, IngramSpark and a conventional printer starts here: with the distinction between making a book, making it obtainable and making anybody want it. These are separate jobs. Publishing has helpfully arranged for all three to cost something.

A catalogue entry is not a queue

Distribution gives a book a route to the reader. It does not supply the reader.

A title can be listed in an ordering system without sitting on a shelf. It can sit on a shelf without being prominently displayed. And it can be prominently displayed while shoppers walk past to buy something involving a detective they already know.

IngramSpark describes its distribution process as sending book information to its retail and library network so buyers can order. That is valuable access. It is not a national chain agreeing to stack your debut beside the till. IngramSpark’s distribution FAQs explain what the service actually delivers.

KDP is similarly explicit: joining Expanded Distribution does not guarantee acceptance by distributors or orders from booksellers and libraries. Its own guidance is rather less excitable than the phrase “global distribution” can become in a sales presentation.

Being available in thousands of bookshops is a splendid achievement for a database. The author still needs customers.

For Tony Marturano, this is the point that gets lost. A shop has finite space. Reader requests, sales, publicity and an audience give it reasons to consider a book. The author’s understandable enthusiasm gives it a pleasant conversation.

Before paying for broader reach, establish what reach means. An ordering route? Physical stock? A sales representative pitching the title? Actual promotion? These are different services, even when somebody has put them under one attractive heading.

Why Amazon is often the practical starting point

Tony’s direct experience is mainly with Amazon/KDP and commercial printers. He investigated IngramSpark but did not adopt it.

Amazon’s attraction is the combination. There is a shop, an existing shopping audience, ordering, payment and fulfilment, with KDP supplying the print-on-demand route. You still have to make your particular book worth finding, but at least you are doing it somewhere people already go to buy things.

Selling through your own website can be excellent when you have an audience. Without one, you have built a very agreeable shop at the end of a road nobody uses. The checkout may be immaculate. It will have plenty of time to admire itself.

This is not a declaration that Amazon must be everybody’s entire business. It is a practical starting question: where are your likely buyers already comfortable shopping, and what would make them take an extra journey?

With retailer-fulfilled POD, copies can be produced as orders arrive. The author does not have to finance a large print run, house it and post each copy personally. Order several hundred POD copies for yourself, however, and you have brought the inventory problem home again. The method of manufacture does not make boxes transparent.

IngramSpark, KDP or both: give each one a job

For a book whose immediate print audience shops on Amazon, KDP alone may be a sensible, manageable beginning. A second dashboard needs a reason to exist beyond looking reassuringly international.

IngramSpark becomes worth considering when you want a trade ordering route for booksellers and libraries, or print choices that suit the edition you are making. That is a distribution decision with commercial terms attached, not an upgrade from amateur to proper author.

Using both can also make sense: KDP for direct Amazon supply, IngramSpark for wider trade distribution. IngramSpark’s current FAQs permit that combination with an ISBN you own, provided you do not enrol that print book in KDP’s Expanded Distribution programme. Keep the edition’s identifying details consistent across the services. IngramSpark’s guidance is the starting point for that setup.

A free KDP ISBN is restricted to KDP; it is not a passport you can take to another printer. If both services are part of the plan, settle ISBN ownership before building the edition. KDP explains the distinction in its ISBN guidance.

The unglamorous part of “both” is maintaining both. Files, descriptions, prices and corrections need attention in two places. Use each printer’s specifications and check physical copies from each. One approved upload does not certify a second company’s output.

You can begin with the route that serves the readers you have and add another when there is a reason. Publishing infrastructure is allowed to grow with the business. It need not arrive fully assembled before the first customer.

The discount is real. So are the returns.

Wider distribution introduces another distinction: the price on the back is not the money that comes back to you.

Booksellers need a margin. Printing costs money. The calculation must include the applicable distribution terms and fees, rather than stopping at a handsome retail price and congratulating itself.

Returnability deserves particular attention. Under IngramSpark’s returnable options, unsold books can generate charges to the publisher at the wholesale cost, with handling or shipping charges where applicable. The bill can exceed the author’s earnings on those copies. Some booksellers favour returnable stock, but making a title returnable is accepting exposure, not simply ticking the “please take me seriously” box. IngramSpark explains the options and charges here.

For KDP paperback sales on Amazon, the current royalty calculation uses a 50% or 60% rate depending on list price and marketplace, then deducts printing costs. That is not 60% straight into your pocket, however confidently an old tutorial waves the number about. Use the current royalty guidance and calculator for your actual edition.

Tony recalls circumstances in which a paperback priced around £10.99 left him roughly a pound or so. That is his experience of particular books and costs, not a current universal KDP rate. It illustrates the scale of the problem: a book can look reasonably priced to its buyer while leaving its author very little room to perform miracles.

Calculate what remains. Then ask whether that amount can support the way you intend to sell it.

A lower unit cost can be a more expensive decision

A commercial printer may offer a much better price per copy at a larger quantity. It is tempting arithmetic, especially when the quotation starts rewarding you for adding another thousand.

But a thousand copies bought and a thousand copies sold are not equivalent entries in the business plan.

You pay for the run before all those readers materialise. You need somewhere suitable to keep it. Copies must reach buyers, events or retailers. Packing, transport, damage and the time spent organising everything belong in the comparison too.

The printer’s price gets better at two thousand copies. Your spare room’s opinion is less enthusiastic.

The useful comparison is the total commitment against realistic sales, not simply one printer’s unit price against another’s. A smaller margin on a copy actually sold may be more useful than a magnificent theoretical margin on a book currently supporting a Christmas decoration box.

Tony’s shorthand is the sprint and the marathon. A concentrated burst of known demand can justify stock. A slow, uncertain accumulation of sales often favours POD. Neither is a law of nature; both are better starting points than assuming optimism qualifies as a purchase order.

Sometimes the beautiful book earns its keep

Tony prefers the physical results he has had from traditional commercial printers: richer blacks, colour and detail he likes better, more finishing choices and a more satisfying object in the hand. His preference for a handsome hardback is not particularly secret.

That is a judgement drawn from his own books and suppliers, not proof that every traditionally printed book beats every POD copy. Compare samples of the specification you intend to buy. “Premium” is an adjective; paper, binding and print are things you can inspect.

There are occasions when that better object has a clear job.

Tony’s supernatural-fiction events in old English churches were built around atmosphere: autumn and winter evenings, readings, questions, dry ice, a crew and signing queues, often with standing room only. The audience had come for the author and the experience. A signed physical book belonged to the occasion.

In that setting, stock is serving people who are actually there. A distinctive edition can become part of what they take away, rather than a luxury specification searching for a business case afterwards.

The same reasoning can apply to a confirmed bulk order, a well-supported event or an established direct audience. Size the run around evidence, delivery requirements and what you can comfortably carry if sales disappoint.

Special editions work best as a response to demand. Foil blocking is unlikely to create it from scratch. Though it will make the unsold copies catch the light beautifully.

“File accepted” is not a quality verdict

Whichever route you choose, inspect a physical copy before treating the production job as finished.

The upload checker has not held your book. It has not admired the binding, noticed the cover image disappearing unpleasantly into a crease or wondered why the spine appears to be edging towards the neighbouring paperback.

Tony pays attention to these details: centring, colour, folds, finish and the relationship between the design and the object. Something can be technically acceptable and still look careless.

“File accepted” means the machine can make your book. It has expressed no opinion about your taste.

Design for manufacturing tolerances rather than assuming every cut and fold will land on an infinitely precise line. Use the supplier’s current template, protect important artwork and text, and correct preventable problems. There is a difference between sensible tolerance and giving up because the preview had a green tick.

Screen colour is another trap. A luminous display and ink on paper are doing different jobs. KDP recommends CMYK images for covers and warns that printed colour can vary; conversion alone does not promise a perfect match. Its cover specifications are useful preparation. The physical proof remains the test.

After publication, Tony sometimes buys a normal retail copy at full price, partly for speed in his circumstances and partly to see what a customer receives. That adds packaging and delivery damage to the inspection: scuffed covers and crushed corners do not become less disappointing because the file was immaculate.

It is an additional check, not a substitute for checking the book before launch. Discovering your production standards through the first customer’s review is a particularly public form of quality control.

Corrections are easier when you do not own a thousand mistakes

POD has another practical advantage: corrected files can be submitted for future production without replacing a whole author-funded print run.

That does not mean every copy ordered immediately afterwards will contain the change. Approval, distribution updates and existing stock can complicate the timing. KDP says changes through Expanded Distribution can take up to eight weeks. Check the route’s actual update process rather than declaring the matter settled because an upload completed. KDP’s timelines spell out that delay.

With a conventional run, the wrong version can be sitting in your house in four figures. Correcting the file does very little to persuade those copies to update themselves.

Proof carefully in either case. POD reduces the scale of some mistakes; it is not permission to manufacture them experimentally.

Start with a book you can keep selling

This decision belongs to the wider job of turning a finished manuscript into a publishing business: a professional product, an intelligible route to readers and economics that survive contact with an actual order.

Get a convincing standard edition working first. Choose KDP, IngramSpark, a commercial printer or a combination because each solves a problem you have. Inspect the result. Know what you earn, what you have committed and who is supposed to buy the next copy.

Then expand when the evidence supports it.

Special editions should follow demand. Otherwise, the only person collecting them is you.

There is nothing small-minded about choosing the version of publishing that lets you continue publishing. You can love beautiful books and still prefer not to spend the next five years manoeuvring around them on the way to the washing machine.

By all means plan the premium hardback and the atmospheric launch. But establish that somebody wants the book before ordering enough copies to build the church.