KDP Listing Checker

No, and this is the most important thing to understand about the tool. Our rules are transcribed from Amazon's published KDP help pages, which cover metadata: lengths, prohibited claims, forbidden characters. KDP's actual review also looks at your manuscript, your cover, your category fit, and things Amazon has never documented publicly. A clean report means our checks found nothing — it is a statement about our rules, not a decision by Amazon. Read it the way you would read a spell-checker: useful, and not the same thing as an editor.

Three Amazon help pages: Metadata Guidelines for Books, Make Your Book More Discoverable with Keywords, and Write a Book Description. We wrote every rule ourselves by reading those pages, and each one carries a link to its source and the date we last checked it — visible next to the finding, not buried on another page. Amazon changes its guidelines without announcing it, so a date that has drifted is information you should have.

A problem is something Amazon's guidelines explicitly prohibit, and we can quote the prohibition. A suggestion is our own judgment — repeated words across keyword boxes, shouted capitals, an over-long keyword phrase — where Amazon has said nothing at all. The two are labelled differently and grouped separately, and a suggestion always says in plain words that it is ours. A checker that presents its own guesses as Amazon's rules is worse than no checker, because you would act on it.

Because the field is. A keyword box holds a fixed number of bytes, not characters, and in UTF-8 an ordinary English letter costs one byte while an accented letter costs two and a Japanese or Chinese character costs three. So "café" is five bytes, not four, and a Japanese keyword fills the box roughly three times faster than an English one of the same visible length. Counting characters would quietly under-report exactly the keywords most at risk of being truncated.

Yes, but only twelve tags: <br>, <p>, <b>, <em>, <i>, <u>, <h4> through <h6>, <ol>, <ul>, and <li>. Notably <h1>, <h2>, and <h3> are not supported even though they look like they should be. We check for unsupported tags and for tags you have opened and not closed — one missing </b> bolds everything after it, and KDP will not repair that for you. Remember that the tags themselves count toward the 4,000-character limit.

The rules run on our servers rather than in your browser, and an account is what gates that. The page, the form, the character and byte counters, the FAQ, and the full methodology are all public and work signed out — you can measure your title and description right now without an account. Only the rule check itself needs one.

No. Amazon's metadata guidelines prohibit "Unauthorized reference to a trademarked term", and that is a real and common reason listings get pulled — but answering it needs a search of the trademark register, not a word list. Our Book Trademark Checker does that, and the results panel links to it. We would rather tell you a check is missing than let silence imply we ran it.

No. Your listing is sent to our server, checked against the rules, and the findings come back — nothing is written to a database and nothing is kept after the response. That is also why there is no shareable results link for this tool: a URL carrying 4,000 characters of your unpublished book description is not something we want to create.