How to choose KDP categories and keywords in 2026

In 2026 KDP gives you three store categories, chosen yourself in the dashboard when you set up or edit the book, and seven keyword fields of up to fifty characters each. The old routine — ten categories requested by emailing support — is gone. Pick the three deepest, most accurate shelves, and fill the seven fields with phrases a real reader would actually type.

That is the whole answer. Here is how to do each part well, and why none of it is an afterthought you bolt on at upload.

How many KDP categories do you get, and how do you choose them?

Three, per format, and you choose them yourself. This changed in mid‑2023 and the change is total: Amazon retired the BISAC‑based picker and the email‑support pathway, and now you select from Amazon’s own browse taxonomy — the same hierarchy a reader sees clicking down the left‑hand rail of the store. You set them in the KDP dashboard: open the title on your Bookshelf, choose Edit book details, and scroll to the Categories section. You can change them whenever you like; updates take roughly a day or two to show. There is no longer any way to request a fourth, and no one to email about it (KDP’s category help page is the live reference).

So if a guide tells you to email KDP for ten categories, it is describing a system that no longer exists. Ignore it.

With only three slots, every choice carries more weight than it used to. The instinct is to reach for the biggest, busiest category you qualify for — resist it. A broad shelf buries you among thousands of titles you cannot outrank. Go the other way: pick the deepest accurate sub‑category you honestly belong in. A precise shelf has fewer competitors, an easier path to a visible rank, and — because Amazon maps your store category to the nearest BISAC code behind the scenes — it also hands a sharper classification to the rest of the trade.

To find those shelves, browse the store the way a buyer does. Open a bestseller close to yours, scroll to its product details, and read the category breadcrumb it actually sits in. Do that for five or six comparable titles and the genuinely reachable sub‑categories reveal themselves. Choose accuracy over ambition every time: landing on a wrong‑but‑smaller shelf wins you a misleading badge and the returns that follow it.

How do the seven keyword fields work?

KDP gives you seven separate boxes, numbered one to seven, each holding up to fifty characters including spaces. These are search terms, not shelves: they feed Amazon’s search engine so your book surfaces when someone types a phrase. Treat each field as room for one short phrase — typically two or three words — because that is how real readers search. You do not separate terms with commas, semicolons or pipes; Amazon reads each field as one continuous string, and every comma or quote mark you add simply wastes a character.

The free research tool that beats every paid one is Amazon’s own search box. Set the dropdown to Books, start typing your topic, and watch what autocompletes. Those suggestions come from real buyer searches, and the order roughly tracks popularity — terms near the top are searched more than terms near the bottom. Type the start of a phrase, note what Amazon finishes for you, and you are reading demand straight from the source. Paid tools such as Publisher Rocket estimate volume and competition on top of that, but the search box alone will carry you a long way.

Two rules save the most slots. First, do not repeat words already in your title, subtitle or categories — Amazon indexes those automatically, so spending a field on them is a field thrown away. Second, write phrases, not a word salad: “slow cooker recipes for two” earns its place; “food cooking meals dinner” does not. Spend the seven fields on the specific, multi‑concept phrases your reader would actually speak — reader benefit, sub‑genre, setting, comparable‑author territory, the problem the book solves.

How should you research categories and keywords together?

They are two halves of one job: helping the right reader find the book. Categories decide which shelves you can rank on; keywords decide which searches you appear in. Research them in the same sitting, off the same shortlist of comparable titles.

Build that shortlist first. List ten books a reader would buy instead of yours — same genre, similar tone, similar promise. For each, note the categories it sits in and the phrases its blurb and reviews lean on. Patterns emerge fast: the shelves that keep recurring are your category candidates; the phrases that keep recurring are your keyword candidates. Then test the keyword candidates in Amazon’s search box to see which ones real buyers actually type, and which return a wall of unbeatable bestsellers you would rather avoid.

The aim is honest fit, not gaming. A keyword that wins a click but mis‑sells the book earns you a one‑star review and a return. Match the metadata to what the reader will actually get, and the rank you earn is one you can keep.

When do BISAC codes matter?

BISAC codes are the book trade’s standard subject classifications — the system bookshops and libraries catalogue by. On a KDP‑only ebook you never touch them: Amazon assigns the matching code automatically from the store category you picked. They become your job the moment you go wide or into print — distributing through IngramSpark or another wide distributor, or chasing bookshop and library placement — because those channels classify by BISAC, not by Amazon’s taxonomy. IngramSpark requires at least one code and recommends up to three.

When you go wide, treat BISAC as a parallel task, not a copy‑paste of your Amazon shelves. Choose the deepest applicable code, the same way you chose your deepest Amazon sub‑category; a precise code improves discoverability across the whole trade, a vague one buries you. If you are still deciding whether to buy your own identifiers for that wider reach, our guide on whether you need an ISBN covers the part of metadata that travels with the book everywhere it sells.

KDP categories and keywords: do and don’t

Element	Do	Don’t
Categories (3)	Pick the deepest accurate sub‑category in the dashboard	Email support for more — the old 10‑via‑email route is gone
Category strategy	Choose smaller, reachable shelves you can rank on	Chase the biggest, busiest category you cannot outrank
Keywords (7 fields)	Write one specific 2–3 word phrase per field, up to 50 characters	Stuff single words or duplicate your title and category terms
Formatting keywords	Separate words with spaces only	Add commas, quotes or pipes — each wastes characters
Research	Use Amazon’s search‑box autocomplete and comparable titles	Guess, or copy a rival’s metadata wholesale
Honesty	Match metadata to what the book actually delivers	Win clicks with terms that mis‑sell and trigger returns
BISAC	Choose deep, precise codes when you go wide or into print	Worry about BISAC on a KDP‑only ebook — Amazon assigns it

Why is metadata part of a ready‑to‑publish file?

Here is the shift in thinking that helps most. Categories and keywords are not a form you rush through at the upload screen; they are metadata — the structured information that makes a book findable — and they belong in the same finished package as your validated ePUB and press‑ready PDF. A book that is technically ready to publish but carries vague categories and wasted keyword fields is invisible on a shelf it could have ranked on.

So plan the metadata while you build the files, not after. Decide your three deepest categories, draft your seven phrases against real searches, and — if you are going wide — pick your BISAC codes in the same pass. Done together, they ship as one coherent record rather than a last‑minute guess.

That is how we hand a book over at Librum: alongside files built to pass KDP and IngramSpark first time, you get the metadata sorted so the book is discoverable the day it goes live, with a fixed quote agreed up front — usually within a day. If you would rather your categories, keywords and files arrived finished and working together, see what our production services cover.