Amazon & KDP

KDP categories: the three slots you actually get

Why the ten-category advice you have read is three years out of date, and what to do with three

You get three categories per format. Not ten. Not two plus eight more if you email support. Three, chosen in the KDP dashboard, and there is no way to get more.

This matters because a large share of the advice ranking for this question describes a system Amazon retired in mid-2023. We checked pages published within the last two months that still tell authors they can select up to ten categories per book. If you have read that somewhere and been confused by what the dashboard actually shows you, that is why.

The current rule

Three categories per format, where a format is each separate version you publish. An ebook, a paperback, and a hardcover are three listings, each with its own three slots. Publish all three and you have nine placements in total. If you are also deciding whether each format needs its own identifier, see our guide on what the $125 ISBN fee actually gets you.

Categories are selected from Amazon's own store taxonomy, the same tree a reader clicks down in the store's left-hand rail. Amazon replaced the older BISAC industry-code picker with this in the same 2023 change. Available options are shaped by your primary marketplace and the audience information on the listing, so two books do not always see an identical list.

Changes take roughly 24 to 72 hours to appear on the product page, and there is no cap on how often you change them.

What went away, and the trap it left behind

Before mid-2023, the workflow was two BISAC codes at upload plus a support email requesting up to eight more. Both halves are gone. KDP support does not add categories on request, there is no Author Central backdoor, and there is no escalation path. Authors who try get a template reply pointing back to the dashboard.

Here is the part worth knowing before you touch anything. A book published under the old system, sitting in more than three categories, keeps those categories. It keeps them only until you update the category section. Edit it once and the book drops permanently to three, with no route back to the others.

So if you have an older title in eight categories and you were about to go tidy up its metadata: do not touch the categories unless you have decided the three you want are worth more than the eight you have.

What Amazon does not promise

Three things that most category guides skip, all of which come straight from Amazon's own guidance:

Your choices are not final. Amazon reserves the right to reassign a book it considers miscategorised. Authors do report opening a listing and finding categories they did not pick.

Categories do not guarantee visibility. Amazon is explicit that category rank depends on customer activity. A category slot creates the opportunity to rank; it does not create sales. A book can display rankings in up to three Best Seller category lists.

Some categories cannot be won. The store contains duplicate and so-called ghost categories, paths where a book cannot realistically earn a visible bestseller placement. Spending one of three slots on one is a wasted slot, and Amazon does not label which is which.

Picking three you can actually rank in

With three slots, breadth is the expensive mistake. "Literature & Fiction" puts you behind hundreds of thousands of titles. A specific path such as Mystery, Thriller & Suspense, then Thrillers, then Espionage gives a new book a realistic shot at a visible rank, and a visible rank in a narrow category sells more copies than invisibility in a wide one.

A practical way to find real paths: open the product page of a book genuinely similar to yours, scroll to Best Sellers Rank, and read the exact category paths listed there. Those are live, real, currently-ranking paths, which is better evidence than any category list someone has compiled second-hand. Amazon does not publish its full category tree in one place, which is precisely why competitor pages are the more reliable source.

One filter worth applying to every candidate: would a reader who found your book on that shelf feel the book belonged there? If not, drop it. A category that generates disappointed readers costs you reviews, which cost more than the placement was worth.

The seven keyword fields, and how they interact

Separate from categories, each format gets seven keyword fields of 50 characters each. They do two jobs: match the searches readers type, and help Amazon shelve your book in categories you could not select directly, since some categories only open up when keywords match them.

That is the only real way to extend reach past the three-slot limit, and it is influence rather than control. You request nothing, and nothing is guaranteed.

Amazon's own keyword guidance says to avoid terms already appearing in your title or your categories, since those are indexed already. Spending a slot repeating your own subtitle is the most common way authors waste one.

Fill all seven, but only with terms that describe the book honestly. An empty field is lost reach. An irrelevant one is a wasted field and, at the extreme, a metadata violation.

The short version

Three categories per format, seven keyword fields, no requests, no workarounds. Pick narrow paths you can rank in, read them off real competitor listings rather than compiled lists, and leave older books' legacy categories alone unless you are certain.