Your ads bid on the words in your listing. Fix the words.
Titles, subtitles, descriptions, and the seven KDP keyword slots — drafted, checked against KDP's rules, and lined up with the search terms your campaigns actually pay for. SelfPub Go doesn't write your book. It works on the fields Amazon indexes.
First cohort: 20 authors. 90 days free. No card, no checkout — we're not selling yet.
Description rewrite · draft 3 of 4
A retired lighthouse keeper inherits a journal that should not exist. Each page is dated tomorrow. Cozy mystery for fans of Louise Penny and Anthony Horowitz.
Keywords expanded
cozy mystery · small town · journal · time loop
Title variants
3 generated · 1 pinned
The four fields that decide what your ads can reach.
Title, subtitle, description, keywords. Amazon reads them to decide which searches your book belongs in — and an automatic campaign can never outrun a listing that describes the wrong book.
Titles and subtitles that fit the category
Feed it your blurb and a reference set of books you choose, and it returns candidates ranked on readability, novelty, and how closely they match the naming pattern your category already rewards. Each candidate shows what it's drawing on, so you edit rather than guess.
Descriptions that pass KDP's HTML rules
The rewriter is HTML-aware: it keeps your hooks, tightens the middle, and validates the markup against what KDP accepts before you save. No surprise rejection at upload time because of a stray tag.
Seven keyword slots, chosen with your ad data in the room
Expand a handful of seeds into a long candidate list, deduplicated against the seven slots you already filled and ranked with estimated demand alongside competitor coverage. Pick the final seven and copy them straight into KDP.
The same words your campaigns bid on
Metadata and ads pull from one place. A phrase you put in your subtitle is a phrase you can check as a search term later — and drop if it never converted.
History you can go back through
Recent runs are kept per tool, so you can re-run with the same inputs, compare two versions of a description, and see what you changed. Your text isn't used to train models.
Metadata is an advertising decision, not a copywriting chore.
01
Choose the words
Pick the phrases readers use for a book like yours, with estimated demand next to each one, and put them in the fields Amazon indexes.
02
Let campaigns match on them
Automatic targeting reads your listing. Manual campaigns bid on the same phrases on purpose. Both are working from the words you just chose.
03
Check what they earned
Search-term reports and royalties land in the same app, so a phrase that never converted can be cut from the campaign and rethought in the listing.
What this deliberately doesn't do.
Plenty of tools will offer to write the book for you. We think that's a different product with a different set of risks, and it isn't this one. The boundary is simple: your manuscript is yours, and nothing reaches your Amazon account unless you put it there.
- Write your book, a chapter of it, or a scene in it
- Publish, upload, or change anything in your KDP account for you
- Apply a title or description without you pasting it in yourself
- Promise a category, a ranking, or a sales result from better metadata
Questions about metadata and keywords.
Does SelfPub Go write books?
No, and it won't. These tools work on metadata — the title, subtitle, description, and keyword fields Amazon indexes and your ads bid against. The manuscript is yours and stays that way. If you want a drafting tool, this isn't one.
What does metadata have to do with Amazon Ads?
Everything an automatic campaign can match on comes out of your listing. Your title, subtitle, description, and keyword slots are what Amazon reads to decide which searches your book is relevant to, so a vague subtitle quietly caps what your ads can reach. Fixing the words is usually cheaper than raising the bid.
Will using this get flagged under Amazon's AI content policy?
KDP asks you to disclose AI-generated content in the book itself. These tools don't touch the book — they draft listing fields you review and paste in yourself, the same way you'd use an editor's suggestion. Read KDP's current guidance and answer their questions honestly; we can't answer them for you.
How do the keyword suggestions get ranked?
By estimated demand and how heavily the term is already covered by comparable books. Demand figures are estimates from Demand Radar and are labeled as such — Amazon doesn't publish exact Kindle search volume to anyone.
Can I use it on a backlist I published years ago?
That's the usual case. Old listings tend to carry a description written before you knew your category and keyword slots nobody has revisited since launch — and they're still deciding what your ads can match on today.