How We Recommend Books: Why Mood, Theme & Voice Matter More Than Genre
If you've ever typed a book title into a recommendation site and got back a list that felt completely wrong — same genre, totally different feel — you already understand the problem that More Like This was built to solve. As a book discovery tool UK readers can actually trust, we don't sort by category and call it a day. We look at the things that actually make you love a book: how it's written, what it makes you feel, the pace it moves at, the emotional territory it covers. Genre is just a postcode. It doesn't tell you who lives there.
Here's exactly how our AI thinks when you ask it to find your next read.
The Problem with Genre-Based Recommendations
Genre labels are blunt instruments. "Fantasy" covers everything from Tolkien's slow, mythology-soaked prose to the whip-fast LitRPG chaos of Dungeon Crawler Carl. "Crime thriller" might mean the cold procedural precision of Tana French or the sun-drenched cosy mystery of Richard Osman. Recommending one to fans of the other because they share a shelf label isn't helpful — it's lazy.
Popularity-based tools have a similar blind spot. They surface what's trending, what sold most last week, what has the most Goodreads ratings. That's useful if you want to know what everyone else is reading. It's less useful if you want to know what you'll love next.
What More Like This Actually Looks At
When you search for a book on More Like This, our AI reads your input as a set of signals — not just a title or genre tag. Here's what it's weighing up:
1. Narrative Pace
Is the book you loved a slow burn that rewards patience — long chapters, deep world-building, time spent inside a character's head? Or does it move fast, with short punchy chapters and a plot that barely pauses to breathe? These are completely different reading experiences, and matching them matters. A reader who loves the relentless momentum of Project Hail Mary by Andy Weir probably isn't looking for something meditative and languid next — even if that other book is technically also science fiction.
2. Emotional Tone
What does the book feel like to read? Is it hopeful or bleak? Tense or warm? Funny and self-aware, or earnest and immersive? Emotional tone is one of the strongest predictors of whether a reader will connect with a book — and it's one of the hardest things to capture in a genre label. We try to match not just what a book is about, but how it makes you feel while you're reading it.
3. Prose Style and Voice
Some readers are drawn to crisp, clean prose — every sentence doing exactly what it needs to, nothing more. Others love voice-driven writing that feels almost conversational, like the author is in the room with you. Others still want something lyrical and dense, where the language itself is half the pleasure. These preferences are real and consistent, and they don't map neatly onto genre. We factor in authorial voice when we surface recommendations.
4. Themes and Emotional Territory
What is the book really about, beneath the plot? Grief and identity? Friendship under pressure? Found family in an unlikely place? The ethics of power? Readers often return to the same emotional territory across wildly different genres — a romantasy fan who loves found-family dynamics might connect deeply with a literary novel that has nothing to do with dragons, if the emotional core is the same. We look for those through-lines.
5. Tropes and Reader Expectations
Tropes get a bad reputation, but they're actually incredibly useful data. "Enemies to lovers", "unreliable narrator", "slow-burn tension", "chosen one with a twist" — these are shorthand for specific reader pleasures. Matching tropes, or deliberately subverting them in a way the reader will enjoy, is part of how we get recommendations right. If you searched for Fourth Wing and loved the enemies-to-lovers tension more than the dragon-riding, that changes what we suggest next.
The Three Results — and Why We Give You Three
Every search on More Like This returns three recommendations, and they're not interchangeable:
- Closest Match — the book that mirrors your search most directly. Same pace, similar tone, comparable prose style. If you loved the original, this is the safest bet.
- Second Match — a well-matched alternative that approaches the same appeal from a different angle. Maybe a different setting, a slightly different emotional register, but fundamentally scratching the same itch.
- Wildcard — the unexpected one. A book that shares the deeper appeal of your search — the emotional tone, the thematic obsession, the narrative energy — but might sit in a completely different genre or come from a place you'd never have looked. This is where readers find their new favourites.
The Wildcard is arguably the most valuable result we return. It's the recommendation a well-read friend would make after thinking about it properly — not "you liked that, so here's another one like it", but "you liked that because of this, and here's something else that has exactly that quality".
Real Examples: How This Plays Out
If you search for Project Hail Mary by Andy Weir
The obvious recommendation is another hard sci-fi novel. But what made Project Hail Mary special wasn't just the science — it was the warmth, the optimism, the problem-solving satisfaction, and the unexpectedly moving central relationship. A good recommendation engine looks for books that share those qualities: hopeful, intellectually engaging, emotionally surprising. Our post on what to read after Project Hail Mary walks through exactly this.
If you search for Dungeon Crawler Carl by Matt Dinniman
Genre says: LitRPG fantasy. But the real appeal is the sardonic humour, the relentless pace, the everyman protagonist in a completely absurd situation, and the surprisingly sharp satirical edge underneath it all. That combination exists in unexpected places — and finding them is exactly what we're here for. We've written about this in detail in our Dungeon Crawler Carl recommendations post.
If you search for Fourth Wing by Rebecca Yarros
The genre label is romantasy. But readers are often responding to specific pleasures: the romantic tension that's drawn out over hundreds of pages, the high-stakes world, the found-family dynamics, the competent heroine. Match those, and you might end up recommending something that doesn't have a dragon in sight. See our Fourth Wing recommendations for the full breakdown.
Why This Is Different from Goodreads
Goodreads is brilliant for tracking what you've read and seeing what your friends are reading. Its recommendation engine is less brilliant — it's largely driven by "people who read X also read Y", which is a measure of audience overlap, not book compatibility. The result is a list of popular books in a vague genre neighbourhood, heavily weighted towards whatever is selling well right now.
We're not trying to replace Goodreads. But for the specific question of "what should I read next, based on why I loved this particular book" — we think there's a better way. One that starts with the book, not the crowd.
How to Get the Best Results
The more specific your search, the sharper our recommendations. A search for "fantasy" is a starting point. A search for "fantasy with slow-burn romance, morally grey characters, and a dark academic atmosphere" gives our AI far more to work with. You can search by:
- A specific book title — "books like The Night Circus"
- An author — "something similar to Kazuo Ishiguro"
- A mood or feeling — "cosy mysteries for a rainy Sunday afternoon"
- A combination of themes — "grief, found family, and dry wit"
- A trope — "enemies to lovers with proper slow burn"
The platform is free to use, searches real UK retailer inventory in real time, and returns results you can actually buy — not books that have been out of print since 2003.
Try It Now
If you've finished something brilliant and don't know what to read next, don't scroll through bestseller lists hoping something jumps out. Tell us what you loved — a title, a feeling, a sentence about what made it special — and we'll do the rest. The best book you haven't read yet is probably closer than you think.