How we work

Research that connects books with readers

Kadaxis closes the gap between how publishers describe books and how readers look for them, in retailer search and AI tools. We agree the scope around your list, research each title, and deliver recommendations your team can review and use.

Long tables covered in trays of catalog cards in the Card Division of the Library of Congress, with staff working among them.
People working in the Card Division of the Library of Congress, between 1900 and 1920. Library of Congress.

Start with how readers look

A relevant book can be absent from a retailer search or an AI answer. Understanding the language of the reader's question is the starting point for finding the gap.

Most metadata is written by people who know books. The readers who would buy those books rarely use the same vocabulary as the people who catalogue them. They search using themes, situations, comparable titles, and practical needs that may be missing from standard descriptive copy.

We examine that language alongside live search results and the information available about the book. The resulting work depends on the service: keyword sets for Amazon, evidence and content recommendations for AI discovery, or advice on an AI workflow.

Publisher field romance · contemporary · family life
Reader query cluster small town second chance romance, single parent, found family, low spice
Where description terms sit, from catalogue language to reader search
  1. BISAC
  2. theme
  3. comp
  4. use case
  5. query

How we work

1

Reader-side research

We study how readers in a book's category are searching, what language they are using, and which adjacencies and comparable titles they cluster around. Sources include retailer search data, category-level query patterns, and proprietary research methods refined across hundreds of thousands of titles.

2

Title-specific synthesis

Every title receives its own research, including in large backlist programmes. Recommendations reflect the book's reader needs, category, and competitive set. List-level delivery can be organized around your workflow.

3

Implementation and review

Your team reviews and approves the work. Keyword files fit your existing ONIX workflow; AI discovery reports provide the supporting evidence and recommended content actions. Your team implements approved changes in its systems, with any further review agreed as part of the engagement.

Agree the inputs and the deliverables

For keyword work, start with the title list, relevant ISBNs, and existing metadata. We deliver title-specific keyword recommendations, formatted for your workflow, with supporting research or visibility reporting as agreed.

For AI discovery (GEO), we agree the selected titles, reader questions, and AI services to examine. The report records observed book appearances, competing titles, and sources, then sets out prioritized recommendations with supporting evidence.

Scope, delivery format, and timing are agreed before work begins. Your team retains approval over every proposed metadata or content change.

Credentials

Named BookTech Company of the Year at the FutureBook Awards in 2016.

Contributor to the Book Industry Study Group's 2018 keywords best-practices standard.

Methodology refined across hundreds of thousands of titles, working with publishers from the Big Five to specialty independent presses.

Tell Chris about your publisher, the list you have in mind, or the discovery question your team is considering.

Talk to Chris