To build a KDP catalog, start with one book that solves a recognisable reader problem. Choose that problem, price the work required to serve it, and document permission for every contribution before commissioning a series. A catalog is easier to evaluate later when its usefulness, costs and rights can be explained title by title.
This guide takes you from a rough idea to a book proposition, production budget and rights register. It also shows an established publisher how to organise a catalog for diligence. It does not assume that an Amazon account or selected books can be sold or migrated. Allow several focused research sessions before committing to production; contributor responses and rights checks can take longer.
1. Choose a reader at a particular moment
Replace a broad topic such as “gardening” with a situation: a renter planning a first container garden on a shaded balcony. Write what that person needs to decide, what constraints matter and what a useful finished book would let them do.
For this hypothetical project, the result is a feasible planting plan, shopping list and weekly care routine. It excludes commercial horticulture, structural balcony assessments and promises of a particular harvest. Those exclusions help you recognise when the proposed book needs expertise you do not have.
Find evidence in reader questions, library catalogs, comparable book descriptions and public reviews. Keep a small research sheet with the source, date, question, context and your interpretation. Paraphrase patterns; do not copy another author's explanations, illustrations or organising structure. A complaint that a competing book lacks local climate information is a lead to investigate, not permission to reproduce its other chapters.
Look for disagreement as well as repetition. One reader may want an illustrated beginner workbook while another wants a compact reference. Combining both without a clear priority can produce an expensive book that satisfies neither. Choose a primary reader and record the readers the first edition will not serve.
2. Test the proposition before buying a full manuscript
Write this sentence: “For [reader] facing [situation], this book helps produce [specific result] through [method], unlike the available alternatives that leave [documented gap].” Remove any superiority claim you cannot substantiate.
The balcony example becomes: “For renters with shaded balconies, this workbook helps plan a small container garden through light observations, container selection and a seasonal care sheet.” That is a production brief. It is not proof that readers will buy it.
Show a proposed contents page and one original sample exercise to potential readers through channels where such requests are welcome. Ask them to explain the task in their own words, identify confusing instructions and say what they use today. Avoid asking only whether they “like the idea.” Praise is easier to obtain than evidence that an exercise works.
Keep the sample modest. You are testing whether the reader understands the promised result, not producing the entire book for free. Do not describe this feedback as a sales test unless people actually purchased under a clear offer. If readers repeatedly need information outside your scope, narrow the promise or recruit a qualified contributor before proceeding.
3. Complete a production brief
A useful brief connects each chapter to a reader action and an acceptance check. In the example, a light-observation chapter ends with a completed seven-day observation sheet; a container chapter ends with a selection justified by space and watering constraints. “Twenty pages about pots” is less helpful to a writer than an observable reader outcome.
Include format, expected illustrations, accessibility needs, reference style, editable deliverables and revision rounds. Specify who checks technical facts and who resolves conflicting advice. A copy editor cannot automatically validate specialist claims, and a subject specialist may not catch navigation or layout problems.
Use the following completed example as a model rather than a universal specification:
| Brief field | Hypothetical decision |
|---|---|
| Reader | First-time renter gardener with limited balcony light |
| Core result | A small planting and care plan matched to observed conditions |
| Evidence required | Specialist review of plant-care statements and original exercise testing |
| Deliverables | Editable text, original diagrams, reference list and layout source |
| Acceptance | Sample reader completes the planning sheet without unexplained steps |
| Exclusions | Structural advice, commercial yields and guaranteed growing outcomes |
This is the point to choose whether the idea warrants a book. A short worksheet or article may serve the reader better. Expanding a thin idea into multiple interchangeable titles creates more production and maintenance work without establishing additional value.
Test a chapter before scaling the catalog
Choose one practical reader task and write a short chapter that enables it. Supply the required inputs, a worked example, an exercise and a way for the reader to check their result. In a hypothetical budgeting chapter, the worksheet must explain the units and distinguish cash timing from profit; a polished cover does not establish that the lesson works.
Ask a target reader to use the chapter without coaching. Separate observations about understanding, navigation and accessibility from copyediting comments and discovery metrics. A friend saying the book is good is not a completed exercise or evidence of paid demand.
Preserve the task in the intended format. A worksheet that requires a wide table or external file needs an appropriate digital or print path and clear instructions. Use KDP's preview guidance for format inspection, and the title quality check for a versioned repair log.
Keep the catalog's rights and account questions separate from manuscript quality. KDP's account guidance does not permit moving individual books between accounts. A title-level earnings sheet cannot establish that a proposed catalog sale has a supported transfer route.
4. Budget cash and owner work separately
Ask contributors for scoped quotes instead of adopting generic per-book costs. Include research, writing, technical checking, editing, illustration, cover, formatting, proof copies, shipping and revisions. Record the currency and whether a quote includes taxes and editable files.
Here is a hypothetical budget, not a market benchmark: writing $900, editing $450, diagrams $250, cover and layout $400, and proofs plus shipping $100. Cash production cost is $2,100. A $300 contingency raises the cash envelope to $2,400. If the publisher spends 30 hours valued internally at $30, the project also consumes $900 of owner time; total planned resources are $3,300.
Those figures do not predict profit or sales. To examine recovery, divide the relevant cost by a realistic per-copy contribution derived from your chosen format, territory and current KDP calculations. Define contribution here as royalty per customer copy after platform deductions, less any additional variable cost you pay per copy; do not use the retail price. At an assumed contribution of $3, recovering $2,400 requires 800 copies before other ongoing expenses. Change the contribution to $2 and the same cash cost requires 1,200 copies. Both are arithmetic scenarios, not sales forecasts.
Keep production investment separate from ongoing advertising, software, updates and management. Later royalties must be reconciled with those expenses before anyone calls the catalog profitable. Do not count a proof purchase as customer demand.
5. Build a rights register that survives personnel changes
Create one row for every text, photograph, diagram, font, cover element and other contribution. Record the creator, source file, written permission or agreement, permitted formats and territories, attribution requirements, expiry and whether the rights can support a future business transaction.
If you are considering merchandise as well as books, review print-on-demand royalties, design rights and transfer routes as a different programme model; do not assume that KDP rules or royalty definitions apply to it.
“Paid designer” is not a rights record. Ask what rights the payment actually purchases and whether third-party elements have separate restrictions. Keep the original license and invoice with the asset, rather than relying on a screenshot of a marketplace thumbnail. Where the proposed use is unclear, replace the asset or obtain qualified advice before publication.
Amazon places responsibility for content compliance on the publisher, including rights and customer experience. Its current guidance also addresses AI-generated content. Check the applicable disclosure requirements against the actual production workflow rather than treating AI assistance as proof of ownership or quality. KDP content guidelines.
For the balcony book, an unresolved stock-photo license should prevent that photo reaching the print file. It need not prevent editing unrelated chapters. Assign a named owner and a deadline to resolve the permission; if neither a usable license nor a replacement exists at file lock, remove the image and repair the surrounding explanation.
Review the actual programme agreement before assignment
Check the current KDP Terms and Conditions against the proposed transaction. At the November 17, 2026 public-document review, KDP's assignment language includes a sale-of-all-or-substantially-all-assets exception with notice conditions. The agreement describes notice within ten business days after the assignment; that is not a general permission to sell an account or a complete account-transfer procedure.
Keep a separate row for the catalogue rights, the agreement, account administration, tax/payment details and any provider confirmation. Have the actual transaction and notice requirements reviewed before signing. Do not copy another Amazon programme's rule: the Associates agreement and Merch agreement use different assignment language and conditions. A buyer receiving files and passwords has not thereby established the right to distribute each title or receive its royalties.
6. Operate a catalog with title-level evidence
After a release, maintain an inventory with title identifiers, formats, publication dates, contributor files, rights, current editions and recurring work. Add monthly royalties, refunds or adjustments where reported, promotion spend and other actual expenses. Preserve export dates and reporting periods so another person can reproduce a calculation.
Separate title economics from catalog totals. A strong total can conceal dependence on one book or a recent promotion. Record which titles require specialist updates and which rely on a contributor who is no longer available. A useful catalog dossier explains who can perform the next revision, what files they need and what it costs.
If you are already considering an exit, separate the proposed business assets from personal Amazon services and credentials. Specify whether the proposal concerns a whole publishing business, a particular rights package or selected titles. These are different scopes; do not present them as interchangeable routes.
KDP explicitly says it cannot move specific books between accounts. Its merge process also excludes historical reports and active advertising, among other records; download reports before any approved merge because old-account report access ends. KDP documents account management and merging, with limits and support involvement. That documentation is not blanket approval for a sale or a partial-catalog migration. Put the exact proposed structure and account scope to KDP and retain its response before making transfer promises. Manage your KDP account.
For a proposed transaction, add three questions to the dossier: Which titles and formats are included? Which contributor permissions cover the proposed use and which need further consent? What exact account process has KDP confirmed for this scope? Record each answer, supporting document and unresolved action. An unanswered transfer question remains a transaction blocker even if the books are profitable.
7. Make the proceed, revise or stop decision
Proceed to the first-title quality check when the reader result is specific, the production envelope is affordable and material rights are documented. Revise when reader feedback exposes a fixable gap. Stop when the proposition depends on borrowed content, unqualified specialist advice or cash you cannot afford to tie up.
For an established catalog, prepare a question list for each unresolved permission, concentration or cost issue. Do not substitute an assumed valuation multiple for that work. Once title-level operation is credible, the digital-asset business plan can help connect maintenance, cash needs and a possible future exit.
What should be finished before commissioning?
Finish the proposition, reader evidence sheet, chapter acceptance criteria, cash budget and initial rights register. Open items should have owners and resolution dates. A large topic list alone is not a commissioning brief.
Does a useful catalog guarantee a sale?
No. Reader value and clean records support evaluation, but buyer demand, provider permissions and transaction terms remain separate decisions. Build a publishing operation you can responsibly run while those questions remain open.

