How to write a review that actually helps the author

Reviewing is a small act of generosity that most of us do badly. Not out of malice — out of habit. We reach for a verdict because a verdict is easy: stars, a thumbs up, a sentence that sorts the book into good or bad. The problem is that a verdict helps nobody. The author cannot use it, and the next reader cannot tell whether your taste resembles theirs.

Separate "not for me" from "not good"

This single distinction improves almost every review ever written. A slow literary novel is not a broken thriller. A romance that ends happily is not predictable, it is keeping the genre promise that made someone pick it up.

So before you write anything, ask: am I describing a flaw, or describing myself? Both are worth saying — but say which one it is. "Too slow for me; I wanted plot and this is a book about a marriage" is genuinely useful to a reader deciding, and costs the author nothing. "Nothing happens" is just a stranger being annoyed.

The shape that works

Three parts, and you can do it in four sentences.

What kind of book this is. Mood, pace, what it is trying to do. This is the part most reviews leave out and the part other readers need most.

What worked, specifically. Not "good writing" — which sentence, which scene, which character choice. Specificity is the whole gift here. "The mother never explains herself and I believed her completely" tells the author which instinct to keep. "Great book" tells them nothing, no matter how kindly it is meant.

Who it is for. End by pointing at the person who should read it. That turns your review from a verdict into a recommendation, which is the thing that actually moves books between people — the same principle as recommending a book to a friend.

How to disagree without trashing

You are allowed to dislike things. Strong reactions are part of reading honestly, and a culture where everything gets four stars is useless to everyone. The craft is in aiming it.

Criticise the book, not the writer. "The middle sags" is about a manuscript. "The writer is lazy" is about a person, and you do not know that person. One of these can be acted on.

Say what you wanted instead. A complaint with a direction in it becomes information. "I never understood why she stayed" gives a writer somewhere to go; "bad characters" does not.

Scale your tone to the writer. A chapter-by-chapter serial by someone publishing their first story is not a novel from a major publisher, and the same review does very different damage. This is not a request to lie — it is a request to notice who is reading. Writers on the receiving end of this will recognise the other side of it in how to handle negative comments on your writing.

Skip the review you do not owe. If you gave up fifteen pages in, you have not read the book, and "DNF, one star" is a rating of your own patience. Keep it for yourself instead — there is a good case for keeping a did-not-finish list privately.

Chapter comments are a different instrument

Serialised books come with a problem reviews do not have: people are reading at different speeds, in the same place, at the same time. A comment on chapter nine is read by someone who has just finished chapter nine — and by someone arriving three months later who is on chapter two.

So the etiquette is simple and worth taking seriously. Never reference anything that happens after the chapter you are commenting on. Not a hint, not a knowing wink, not "wait until you see what she does." Predicting is fine; confirming is not. If you have read ahead, comment where you are, not where you have been.

And pitch the comment to the moment. Mid-story is not the time for structural notes on the first act — the author cannot rewrite what is already published, so a critique of chapter one delivered at chapter thirty is pure cost with no use. Save the craft feedback for the end, or for a message. The general rule for talking about books in mixed company is laid out in how to talk about books without spoilers.

Write the review you would want

The best test is selfish. Imagine you wrote the book: which of your reviews would you reread at a bad moment, and which would you learn something from? They are often not the same review, and the good ones are rarely the kindest — they are the most specific.

And remember that the review is one of several ways to be useful. Finishing someone's book, telling one friend about it, leaving a comment on the chapter that got you — those move a writer's week more than a rating does. We go through the rest of them in supporting writers you love.

We make Arwy, an Android app where books are published a chapter at a time; you can comment on a chapter, or on a single paragraph inside it, which tends to produce the specific kind of response this whole piece is arguing for. If what you want is a library catalogue, shelves and detailed rating statistics, there are dedicated tools for that and they do it better than we do.

---