Methodology

We use book metadata, public interest signals, and human editorial checks to build pages that are easy to read and actually help you decide.

Inputs

• Publisher and catalog metadata (titles, contributors, categories, publication details).
• Public metadata signals used to identify demand and relevance patterns.
• Text fields such as descriptions, topics, and list intent language.

How Lists Are Assembled

• Lists are ranked with quality and relevance signals, then cleaned for reader-friendly naming.
• Books are selected for fit strength with the list intent and internal-link value.
• Low-utility taxonomy artifacts are filtered from public-facing labels.

Corrections Policy

If you spot a factual issue, broken page, or poor fit recommendation, contact us and include the URL. We review corrections requests on a rolling basis and prioritize factual fixes and user-safety issues first.

Freshness Policy

We refresh trend-facing pages on a rolling cadence and log what changed in Updates. Last release refresh: 2026-02-27.

Current corpus snapshot: 20008 books, 189 active lists, 0 topics, and 50 trope hubs.