AuthorAZ
Metadata

The Author Back Office: What to Track Beyond the Book

Every published book generates paperwork the writing advice never mentions. The full inventory of what you have to track, and where it belongs.

Alexandru Filip
Aug 10, 2026
8 min read

There is an enormous amount of writing advice about writing. There is almost none about the second job that starts the day you publish.

Call it the back office. It is the paperwork of being an author: the identifiers, the descriptions, the versions, the links, the receipts, the proof. None of it is creative work. All of it is load-bearing. And most indie authors discover it the way you discover a step in the dark — by missing it.

The specific failure is rarely dramatic. It's the paperback that went live with the ebook's description. It's the four covers named cover-final-v3-REAL.jpg and no memory of which one is on the retailer page. It's spending twenty minutes finding an ISBN you have definitely owned for three years.

Individually, trivial. Collectively, they are the reason book five feels harder to launch than book one, despite four books' worth of experience.

Here is the full inventory.


1. Identity layer

The layer above books. Most authors don't formalise it, then hit a wall the first time they publish under a second name.

Per pen name, you need on record:

  • Legal name and the pen name it maps to
  • Author bio in at least three lengths — 50, 100 and 250 words. You will be asked for all three, always at short notice
  • Author photo or brand mark, in the resolution the requester wants rather than the one you have
  • Website and social handles for that identity
  • Brand colors and fonts, so graphics from 2024 and 2026 look like the same author
  • Which retailer accounts and author-central profiles belong to it

The trap: pen names are not just a display name. They are separate brands with separate readers, separate lists and, occasionally, separate tax and payment arrangements. Treating them as a field on a book record rather than a top-level entity is the mistake that makes book three under the second name genuinely painful.

2. Book record layer

The core. Everything below repeats per title.

Descriptive

  • Title, subtitle, series name, series number
  • Blurb or description — and the HTML version, which is different and which you will re-paste more often than you expect
  • Short pitch, one line, for newsletters and swaps
  • Comp titles and the positioning statement you wrote once and forgot

Commercial

  • BISAC categories, and the retailer categories you actually selected (they aren't the same list)
  • Keywords — Amazon KDP allows seven keyword fields, and the ones you chose are worth keeping because you'll want to test against them later
  • Price by market and by format
  • Release date, preorder date, and the dates of any promotions run

Structural

  • Page count, word count, trim size, and file specifications per format
  • Rights and territories, if you've licensed anything
  • Edition history — the thing nobody records and everybody needs

3. Identifier layer

Boring, small, and the single most retrieved category of information in an author's life.

  • ISBN per format. Each format needs its own: ebook, paperback, hardcover, audiobook. Same book, different products. (See: How many books can one ISBN cover?)
  • ASIN for anything on Amazon
  • Imprint name, if you have one, and which ISBNs it owns
  • Copyright registration number and date, where applicable
  • Barcodes, if you generated them separately

These live in your head badly and on paper well. There is no version of an author career where you stop needing to look up an ISBN.

4. Asset layer

Where the version-control disaster actually happens.

  • Cover files, per format and per revision — ebook, paperback wrap, hardcover wrap, audiobook square. These have different dimensions and different bleed requirements, and they are not interchangeable
  • Interior files, per format, with the version that was actually uploaded clearly marked
  • The manuscript itself, and which draft corresponds to which published edition
  • Marketing graphics, quote cards, ad creative
  • Any 3D mockups or promotional renders

The single most useful discipline here is not folder structure. It's a "current" flag. For every asset, one file is the live one. Everything else is history. Authors don't usually lose files; they lose certainty about which file is real, which is worse, because you can't tell it's happened until you've uploaded the wrong one.

5. Distribution layer

  • Which platforms each title is on, and in which formats
  • Account and dashboard locations — and, critically, which email address each account uses, because after four years there will be three
  • Buy links per store per territory. Amazon alone will generate a dozen
  • Universal book links, if you use them
  • Exclusivity status and expiry dates. If a title is in an exclusive program, the date it auto-renews matters and is easy to miss
  • Distribution partner, per title, per format

6. Proof layer

Everything that demonstrates the book is worth reading. Chronically under-organized because it arrives unpredictably.

  • Editorial quotes, with attribution and the date they were given
  • Reader reviews worth quoting — captured at the time, because reviews get edited and deleted
  • Awards, shortlistings, longlistings
  • Press mentions, interviews, podcast appearances, with links
  • Blurbs from other authors, and permission to use them

The reason to capture these at the moment they appear rather than when you need them: when you need them, you need them in an hour, for a media kit or a submission, and the review you remember from eighteen months ago is not findable by search.

7. Money layer

  • Sales and royalty statements, per platform, per period
  • Production costs per title — editing, cover, formatting, proofing
  • Marketing spend, ideally attributable to a title rather than a month
  • Tax documents, VAT/GST registration where relevant, withholding forms
  • The per-title P&L, which is the only number that tells you whether a book worked

Covered properly in The P&L: How Money Actually Works.

8. Operational layer

  • Launch checklists, and the notes about what went wrong last time
  • Contacts: editors, cover designers, narrators, proofreaders, formatters — with rates and turnaround times
  • Contracts and agreements
  • Newsletter and back-matter blocks, reused across books
  • The running list of things to fix in the next edition

Where each layer should live

The inventory above is not an argument for putting everything in one place. Some of it genuinely belongs in the cloud.

LayerSensible homeWhy
Identity, book records, identifiersStructured, local, one source of truthRetrieved constantly, small, and yours
AssetsLocal, with encrypted off-site backupLarge, versioned, irreplaceable
Distribution linksStructured, localCopy-pasted weekly
ProofStructured, localNeeded at short notice, in a hurry
MoneyWherever your accountant can reach itSomeone else needs access
OperationalWherever you already workCollaborative by nature

The distinction that matters is not cloud versus local. It's who else needs to reach this. Your accountant needs the royalty statements. Nobody needs your unpublished manuscript, your unreleased cover, or the pen name you haven't announced.


The realistic starting point

You do not need to build all eight layers this week. In order of return:

  1. Identifiers. One page. Every ISBN and ASIN you own, per format. This takes forty minutes and you will use it for the rest of your career.
  2. Current assets. For each book, mark which cover file and which interior file is the live one. Delete nothing; just flag the current.
  3. Book records. Title, blurb, HTML blurb, categories, keywords, price, dates.
  4. Proof. Start capturing from today rather than reconstructing backwards.

The rest fills in as it becomes annoying enough to matter.


Where AuthorAZ fits

AuthorAZ is made by the author of this site.

The back office is exactly what AuthorAZ was built for — layers one through five and seven, kept on the device rather than distributed across a note app, three spreadsheets and a browser folder. Books, editions, formats, ISBNs and ASINs, blurbs and HTML blurbs, keywords, covers per format, buy links per store, pen names as first-class identities with their own branding, and quote and review banks. One-tap copy for the fields you paste constantly, and encrypted export for your own backups.

It is not where you write and not where you format. It is where the operation lives once the writing is done.

Plenty of authors run all of this in a spreadsheet, and a well-kept spreadsheet is genuinely fine. The failure mode of the spreadsheet is that it doesn't hold the assets, so you end up with the records in one place and the files in another.


Key Takeaways

Eight layers: identity, book records, identifiers, assets, distribution, proof, money, operations.

Authors rarely lose files; they lose certainty about which file is current.

Start with identifiers — forty minutes that pays back for a whole career.

Free · No account · 79 lessons

Read the free Publishing Course

The whole Publishing Manual, free and online: checklists, templates and decision trees for every stage, with platform figures verified against the Publishing Database.