There’s a particular sort of digital garden in which the gatekeeper offers a handsome key, a polished sign, and a very persuasive explanation of why the other gates are overrated.
The key is genuinely useful.
It opens a subscription programme, a set of promotional tools, and a dashboard reporting sales and pages read within a day or two.
It also comes attached to a term sheet.
Which is where a surprising number of conversations about going exclusive turn into horticulture by slogan. Amazon on one side, everyone else on the other, and a library somewhere behind a hedge looking reproachful.
That is not the garden’s actual layout.
Enrolment runs in fixed ninety-day terms.
It concerns the Kindle ebook edition. Not the print or audio editions of the same title.
The current terms also carve out an exception for libraries.
And a retail ebook sale, a subscription read, and a library loan are three different commercial events with three different pieces of paperwork.
So your title can show impressive traffic through one gate while another acquisition system cannot see it at all.
The useful question isn’t whether wide is nobler, or exclusivity more commercial.
That’s the binary preferred by people who have never had to reconcile a royalty dashboard against a library order.
Which reader route disappears, and what is the remaining route actually paying for?
Here is how to inspect the gates.
Read the deed, not the ivy
The diagnostic comes before the strategy deck.
What format and term are covered?
If the answer is the book is exclusive, the answer is not yet fit for purpose.
Exclusivity here restricts the Kindle ebook edition during an active ninety-day term.
Print and audio may travel elsewhere.
A paperback can sit in an independent bookshop while your ebook is enrolled.
An audiobook can take a separate path entirely.
Neither arrangement breaks an ebook-only restriction.
That distinction matters, because your title gets treated as a single plant when it is a small estate with several entrances.
The file makes the same point.
The EPUB that begins as XHTML, CSS, a package document, and a navigation table of contents is not the Kindle file delivered to your customer.
Amazon converts your submitted EPUB into its own rendering format and wraps it in Amazon DRM.
Apple wraps its EPUBs in FairPlay.
Other retailers use Adobe’s system, or increasingly an open successor.
DRM is a lock on a delivered copy.
Exclusivity is a covenant about where a particular edition may be distributed.
Confusing the two is how a team ends up debating padlocks while the gate itself stands open, or shut, somewhere else entirely.
There’s a stereotype that an exclusive title belongs to Amazon the way a house cat belongs to a warm patch of sun.
It does not.
It has agreed to a format-specific arrangement with a term.
Your garden may still have a print gate, an audio gate, and, under the current terms, a library gate.
The test isn’t philosophical.
Read the deed for the edition you’re actually discussing.
Count the closed paths
Once you’ve named the format, inventory the routes rather than admiring the busiest one.
An ebook sold wide reaches multiple retailer dashboards, and you reconcile each platform’s reporting separately, because no unified statement exists across retailers.
An enrolled ebook gains the subscription programme and the promotional tools, and cannot test ordinary wide retail during the term.
Nor does cancelling mid-term restore it. The current term has to end.
That’s a closed path.
It is not a moral failure, and it is not proof that a wide release has more readers.
It’s a loss of simultaneous availability.
Which means a category rank can become noisily beautiful while every other storefront, and any retailer-dependent discovery elsewhere, has no retail object to acquire.
The platform permits up to ten category selections, and a title can rank number one in a narrow category on fewer sales than a broad one would demand.
Your rank is a weather vane in one part of the garden.
Not an aerial survey of the estate.
And libraries need their own count.
The library route diverges after distribution, through a lending platform. You supply the licensing terms: cost per licence, duration or loan cap, metered or one-copy-one-user treatment. The library decides how many licences to acquire.
Which means exclusivity is not accurately described as a universal library ban.
It does mean you check the current permitted path, rather than smuggling a retail assumption into a lending decision.
A library buyer is not a reader purchasing an ebook at midnight.
It’s an institution acquiring capacity.
Price the gate
The seventy per cent figure has the tidy shine of a brass gate latch.
It is also very good at hiding the garden behind it.
That rate applies only inside a price band, in designated territories, and it’s reduced by a delivery cost that varies with file size.
The lower rate applies at any list price, in any territory, with no delivery deduction.
That is a tier rule.
It is not an exclusivity rule.
A thirteen-dollar ebook does not become seventy per cent simply because its author wants the handsome latch. Nor does a title sold outside an eligible territory.
So the correct calculation is never seventy per cent versus trade, still less seventy per cent versus library.
It’s price band, territory, delivery cost, and the channel whose revenue event you are counting.
For a trade-published ebook, the familiar rate is a quarter of your publisher’s net receipts.
Under agency pricing, the publisher sets the retail price and the retailer takes a commission, leaving the rest as the publisher’s receipt before your author’s share is calculated.
Under wholesale, the publisher sells at a discount and the retailer sets the customer price.
The same sale can be planted under either arrangement.
The question is who gets to prune the consumer price.
Not which system has better taste in roses.
Fountain, or irrigation
The subscription programme looks, from a distance, like a purchase with more reader enthusiasm in it.
It is not a sale at all.
Enrolled titles are paid from a monthly global fund in proportion to a standardised page count, rather than per borrow.
The fund runs to tens of millions of dollars a month across authors and territories, and the reader-level count is capped per title per customer per read-through.
And the page rate arrives after the month closes.
Third-party trackers have put it somewhere near half a cent for years now.
It is not a rate anybody promises you before the water starts running.
So a three-hundred-page completed read nets you something like a dollar and a third.
A five-dollar direct sale at the higher royalty tier produces closer to three and a half, before delivery deductions.
Neither number hands down a verdict.
A serial author with fast read-through is buying a different kind of traffic from a single-title author seeking direct sales across storefronts.
The useful distinction is subscription consumption versus retail purchase.
Pages read can make a platform rank and an income line move together.
That tells you nothing about how many buyers were unavailable to a non-Amazon storefront, or which readers would have arrived through a library catalogue instead.
Walk two novellas down different paths
Consider two romance novellas with equal editorial care, comparable covers, and the same talent for ending a chapter at precisely the moment a sensible person should make tea.
The first is the opening gate in a serial read-through. Its next title is already visible.
The author takes one ninety-day exclusive term, qualifies for the subscription programme, and uses the promotional tools available inside it.
The wager is not that subscription pays more per individual reader.
The arithmetic above suggests the opposite.
The wager is that a subscription reader who finishes one compact instalment moves through the garden into the next, and that the series amortises attention across the backlist.
The second novella is designed for a wide library path.
Its retailer availability stays simultaneous across storefronts, and its library edition goes through the separate lending route, where an institution buys capacity on terms that expire.
The library decides the licence quantity. When all licensed copies are out, the app builds a hold queue.
That is not a discounted retail sale wearing sensible shoes.
It’s institutional acquisition, with a queue as its demand signal.
Wide is always better is the stereotype, usually delivered with the confidence of somebody who has never watched a romance series earn its keep through sequential reading.
It is false.
Exclusive is always better is its identical twin in a branded polo shirt.
Also false.
The first novella fits when repeat subscription consumption and serial visibility are the actual engine.
The second fits when simultaneous storefront access and library acquisition are part of its market position.
The difference is fit.
Ask your library buyer
Library lending is where gated-garden language most often stops being useful.
A library does not buy a file and then make endless free copies in a little moonlit shed.
It buys licences.
A metered licence expires after a stated period or a loan count. A one-copy-one-user licence serves one borrower at a time until removal. The time-limited copy returns to the pool automatically at the end of a loan, and the file is never replicated per borrower.
Which makes your author’s revenue event unusually crisp.
The publisher is paid once for the licence, and the author’s royalty is typically a share of that licence revenue rather than a payment per checkout.
One licence may generate dozens of loans before it expires.
The queue, meanwhile, reflects how many licences the library chose to buy.
Not a royalty meter ticking behind each reader’s ear.
None of which makes libraries a prestige annex for books that have outgrown retail.
Nor does it make a library licence a better sale.
It describes a separate acquisition system, with a buyer, a budget, terms, and a capacity decision.
A romance novella with club circulation, local demand, or backlist legs may have every reason to keep that gate conspicuous.
Another may properly put its weight behind the subscription route.
Your library question is not whether the author supports libraries in the abstract.
It is whether lending capacity is part of your book’s reader strategy.
Change one trellis at a time
The most expensive way to learn from a distribution experiment is to redecorate the whole garden, change the gates, and then declare the begonias decisive.
A ninety-day term gives you a natural timed window.
It does not turn every change inside that window into evidence.
So set your exit metric before enrolment.
It might compare pages read and direct sales against the retail availability forgone. It might treat library acquisition as a separately observed route under the stated exception.
Then hold the other packaging choices still.
Change one variable. Price, cover treatment, description, or category placement.
A price drop, a new cover, a revised description, a fresh release, and enrolment all at once produce a dashboard full of confetti and no inference whatever.
Which matters particularly because your sales and pages-read data arrive with a day or two of lag, while payment generally arrives about two months after the sales month closes, and longer through expanded distribution.
Rank is immediate enough to become emotionally persuasive.
Cash and channel evidence are slower, less charming, and far better suited to deciding what the next term should do.
Leave a gate open on purpose
The ambiguous case is neither the single-title wide purist nor the serial author whose readers plainly live inside the subscription fence.
It is the ebook with credible read-through and a credible need for more than one acquisition route.
Give that book one permitted ninety-day term.
Name the exit metric before the first page is borrowed. Sustained read-through relative to direct sales, perhaps. Evidence that the library route is doing work the retail route cannot. Or the absence of either.
Do not use a category rank alone.
A number one in a narrow bed is still a number one.
It is simply not the whole garden.
At term end the choice is practical.
Renew if the subscription path is demonstrably serving the book’s actual release pattern.
Leave if the forgone storefront or library opportunity is the route the book needs.
Agency, wholesale, subscription, lending, DRM, and retailer conversion do not become one system because you discussed them all under the word ebook.
One crowded path can be a splendid thing.
It can also be the sign at the garden wall telling every other visitor to go home.