Book Recommendation Engine

Readers struggle to find their next book to read. A recommendation tool trained on millions of reviews learns your taste and suggests books based on what you've already read. Target: book lovers, Goodreads users, and avid readers.

Goodreads has 150 million users and Amazon owns it, yet the recommendation quality remains notoriously poor — that gap has existed for years and still frustrates readers, which is why indie alternatives keep getting built. The closest incumbent is StoryGraph, which has carved out real traction specifically on the promise of better recommendations than Goodreads, meaning you're not competing against a complacent giant but against a focused, well-regarded startup with a loyal user base. The $500–3k/mo revenue band is honest but sobering — readers are notoriously price-sensitive, freemium conversion is slow, and the audience skews toward people who want free tools, making it hard to justify the infrastructure and data costs of running ML pipelines at that revenue ceiling. The most likely failure mode is the cold-start problem compounded by StoryGraph's head start: without a large existing user graph, early recommendations are weak, weak recommendations don't retain users, and you never build the data flywheel the product depends on.