Most book reviews fail in one of two ways. Some retell the story until there is nothing left for a reader to discover. Others announce a verdict — brilliant, overrated, unreadable — and never explain how they got there.
A review that readers trust does something harder. It shows its working. It tells you what kind of book this is, what it sets out to do, how well it manages the attempt, and whether you are likely to enjoy the result. That takes discipline, and a method. Here is one, in seven steps.
1. Read It Twice, and Take Notes the Second Time
The first read is for pleasure, or irritation, or the strange combination of both that usually means a book has done something interesting. Do not stop to analyse. Mark pages if you like, but keep moving.
The second read is the working one. Go back with a notebook open and record observations rather than opinions. “Chapter nine: the funeral is over in six pages” is an observation. “The pacing is bad” is a verdict you have not yet earned. Observations accumulate into an argument; verdicts arrive too early and stay too long.
Note the things that snagged: a scene that felt rushed, a sentence you admired, a character who vanished for two hundred pages. Note the shape of the whole, too — where the book spends its time, what it keeps returning to, what it never quite resolves.
2. Work Out What the Book Is Trying to Be
Judge a book against its own ambitions, not someone else’s. A comic thriller is not a failed literary novel. Ask two questions: what does this book want to do, and what promise does its opening make?
Genres have contracts. A crime novel promises a mystery and some kind of resolution; a family saga promises depth of time; a comic novel promises that the jokes will keep coming. When a book breaks its contract deliberately and well, that is worth saying. When it breaks the contract by accident, that is worth saying too.
3. Judge the Plot Without Spoiling It
You can describe a novel’s setup and still leave plenty to discover. The reliable rule is the first act: you may say what happens in roughly the opening quarter, because a publisher’s blurb does much the same. Beyond that, describe pressure rather than events. “A marriage frays over one long summer” tells a reader what they need. A list of what happens in chapter nineteen does not.
Some spoilers matter more than others. Cut:
- the twist, reversal or revelation, however famous;
- who dies, especially when the book has been building towards it;
- how the central question is answered — including whether it is answered at all;
- the arrival of a character whose appearance is itself a surprise.
Test every detail with one question: does knowing this remove the pleasure of turning the page? If it does, leave it out. “The final third rearranges everything you thought you knew about the family” is not a spoiler. Naming what gets rearranged is.
4. Characters: Look for Pressure, Not Likability
Likability is a poor test. Plenty of great characters are tiresome, and plenty of agreeable ones are dull. Ask instead what each major figure wants, what they do when that want is refused, and whether they change — then ask the harder question of whether the change is earned by what happened, or simply announced in the last chapter.
Notice who carries the book’s ideas rather than its action. Read a page of dialogue aloud. If every character sounds like the same clever person, say so, and give an example.
5. Style: Describe the Mechanism, Not the Mood
“Beautiful prose” tells a reader nothing. Name what the sentences are doing: long, clause-heavy periods; short declaratives that land like taps on a table; a narrator who withholds; dialogue without speech marks; present tense used to keep everything uncomfortably immediate.
Then quote briefly — a phrase or a sentence, never a paragraph — so the reader can see the thing you are describing.
One quoted line, well chosen, tells a reader more about a book’s style than a paragraph of praise.
If you did not get on with the style, be specific about why. “I found it airless” is fine, provided you say which choices made it feel that way. Avoid the easy sneer; it reads as performance, not judgement.
6. Place the Book in Its Context
Context is not name-dropping. It is the comparison that helps a reader decide. “If you liked this, that one does the same thing differently” is useful. “A worthy addition to the tradition” is not. Compare on a specific point: structure, voice, humour, moral seriousness.
Three sources of context are usually enough:
- the author’s earlier work, if a reader might know it;
- two or three comparable books, chosen for a particular quality;
- the circumstances that shape the book — a translation, a historical setting, a genre the author is writing into or against.
Add one honest line about your own position. Readers are rightly sceptical of a critic who pretends to be neutral. If you came to the book cold, or read it in two sittings on a train, say so.
7. Write the Review in Four Moves
By now you have notes, an argument and evidence. The writing should feel like tidying, not inventing.
- Open with what kind of book this is and what kind of time the reader is in for — no plot summary in the first sentence.
- Make the book’s strongest case: what it does well, with one example.
- Give its limits: what it attempts and does not manage, again with an example.
- Close with who should read it, and on what terms.
Keep plot summary to a third of the review at most. Prefer the concrete to the abstract: “the middle section loses its nerve” beats “the novel struggles with structure”. Cut adverbs. Read the draft aloud, then check every name, title and translator credit against the book itself.
The Test of a Trustworthy Review
Before you publish, ask one question: after reading this, would a stranger know whether they want to read the book, and why? If the answer is yes, you have done the job.
Two habits are worth keeping. Write for a reader who has not read the book but might — not for other critics, and not for the author. And assume you may be wrong. A review that makes a clear case while leaving room for a different reading is more persuasive than one that pretends to certainty. Say what you think, show how you got there, and let the reader decide.
Photo: cottonbro studio / Pexels
