Skip to content

Your TBR says more than your bio ever could

The person with 200 books on their Goodreads shelf, a 5-star rating for every Ursula K. Le Guin novel, and a DNF list they are not ashamed of is telling you exactly who they are.

Try the demo

How it works for readers

Connect your shelves. Goodreads, StoryGraph, LibraryThing - we pull your read books, ratings, shelves, and favourite authors. Your reading list stays private. Only taste signals are shared.

The matching engine understands books at multiple levels:

Niche authors hit different

Mainstream
Colleen Hoover
Popularity 85 → NicheWeight 0.15
Niche
Olga Tokarczuk
Popularity 6 → NicheWeight 0.94

Ratings are the sentiment layer. Both having "read" the same book is presence. Both rating it 5 stars is genuine love. If you gave Flights 5 stars and they gave it 2, the overlap vanishes. The algorithm uses the geometric mean of both ratings as a sentiment factor.

What gets scored

Shared authors weighted by Goodreads popularity - obscure literary fiction scores higher than bestsellers
Star ratings on shared books - the sentiment factor filters out shallow overlaps where one person loved it and the other did not
Genre affinity - not just "fiction" but speculative fiction, literary horror, autofiction, translated literature, solarpunk
Books per author - one book is awareness; reading the full bibliography is devotion
Shelf overlap - shared custom shelves ("comfort reads", "made me cry", "annual re-reads") reveal shared values
Shared DNFs and dislikes - authors you both rated below 2 stars, genres you both avoid. A mutual distaste for a hyped book is a real signal
Reading pace - are you both devouring a book a week or savouring one a month? Lifestyle alignment

A five-star match

"You share 6 authors with a combined Goodreads popularity under 10%, your genre profiles overlap 83% on literary fiction and speculative fiction, and you have both rated 9 of the same books within half a star of each other. Your TBR overlap is 22 books."

Not "you both like reading." The actual shelf data.

The algorithm understands readers

It knows the difference between someone who reads 60 books a year across niche genres and someone who reads 5 bestsellers. Both are valid - but they are different lifestyles, and the algorithm weights accordingly.

The deep cuts matter most. If you have both read and loved a book with under 5,000 Goodreads ratings, the NicheWeight for that author approaches 1.0. That single overlap can outweigh 10 shared bestsellers.

Integrations

Goodreads
Read books, ratings, shelves, reviews, TBR, favourite authors
StoryGraph
Reading stats, mood tags, pace, genre breakdowns
LibraryThing
Catalogue, tags, ratings, collections

All opt-in. All revocable. Your reading list is never shown to matches - only the compatibility signals.

Affinity Atlas is in development

No matching is live yet. If you think Goodreads compatibility should be a dating signal - or you have strong opinions about how reading taste should be weighted - get in touch.

Stay in the loop