How to Write a Book Review: The Six-Part Structure (2026)

Quick answer: A book review is your opinion of a book, backed by reasons — not a summary of it. Use six parts: a hook, what the book is, what it's about (without spoilers), what works, what doesn't, and who it's for with a verdict. Keep the plot to no more than a quarter of the review; the rest is your judgement, with examples from the text to support it.

Key takeaways

• A review is an argument, not a summary. The plot exists to support your opinion — it is never the review itself.
• Six parts, in order: hook · what the book is · what it's about (spoiler-free) · what works · what doesn't · who it's for + verdict.
• The one-quarter test: if more than about a quarter of your review is plot, the review has become a book report.
• Reasons, not adjectives. "It was good" is not a review. Why it was good — with a line, a scene or a structural choice — is.
• The spoiler rule: you may name a twist exists; you may not give it away. Write the review a reader can safely read before they buy.
• The same skeleton works for a blog post, a Goodreads/StoryGraph entry and an Instagram caption — only the length changes.

Review vs book report vs summary — the three get confused

Most weak reviews are really summaries: they tell you what happens and stop there. The three forms are not the same thing, and knowing which one you are writing fixes most of the problem.

	Summary	Book report	Book review
Answers	What happens?	What happens, and did I read it?	Should you read it, and why?
Centre of gravity	The plot	The plot + a little context	Your judgement
Opinion?	None	Optional, usually thin	Required, and supported by reasons
Plot share	All of it	Most of it	About a quarter at most
Signed by	—	A student, for a teacher	A reader, for other readers

If your draft could be read aloud as "and then this happened, and then this happened", you have written a summary. The fix is not to delete the plot — it is to make the plot serve a point you are making about the book.

The six-part structure

Use these six blocks in this order. They work whether the review is 150 words or 1,500.

1. The hook — one or two sentences that make a reader want to keep going: a striking detail, a question the book raises, or the exact feeling of finishing it.
2. What the book is — title, author, genre, and the shape of it (a 300-page literary debut; a 12-chapter non-fiction argument; volume two of a trilogy). One or two lines.
3. What it's about — the premise and its central conflict, without spoilers. Enough that a reader knows whether the subject appeals.
4. What works — the strongest part of the book, with evidence: a line of prose, a character choice, the structure, the pacing. This is where the review earns trust.
5. What doesn't — the honest weakness, also with evidence. A review with no flaws reads as an advert, not a review.
6. Who it's for, and the verdict — the reader you would hand it to ("if you liked X", "worth it for the middle section alone"), and a clear closing judgement.

Part by part: what each section must do

1. The hook. Do not open with "This book is about…". Open with the thing that made you react — the sentence you underlined, the question you were left with, the scene you are still thinking about. The hook is a promise that your opinion is worth reading.

2. What the book is. Give the reader the facts fast: title, author, genre, length, and where it sits in the author's work or in its genre. No opinion here yet — just orientation.

3. What it's about. State the premise as the book frames it: who wants what, and what stands in the way. Resist retelling the whole arc. The test is simple: could a reader who has not read the book still follow your opinion afterwards? If you need the ending to make your point, you are about to spoil it — see the spoiler rule below.

4. What works. One clear strength beats five vague compliments. Name it, then show it: the restrained dialogue in chapter four, the way the timeline folds back on itself, the non-fiction argument that survives its own counter-example. Evidence is what separates a review from a rating.

5. What doesn't. Say it plainly and fairly. A slow first third, a subplot that never pays off, a thesis that overreaches. Weakness is not the same as "I didn't enjoy the genre" — criticise the book against its own aims, not against the book you wanted it to be.

6. Who it's for, and the verdict. Finish by placing the book with a reader. "For anyone who liked [comparable title]." "Read it for the essays, skim the middle." Then land the verdict: recommend, recommend with a caveat, or don't. Do not hedge your way out of a judgement.

How much plot is allowed — the one-quarter test

Plot is scaffolding, not the building. A simple check: if you removed every sentence that states what happens, would a review of the book still be left? If not, you have written a plot summary with opinions attached.

Aim for roughly a quarter of the word count on what happens, and three-quarters on what you make of it. So in a 400-word review: about 100 words of premise and events, 300 words of judgement, evidence and verdict. If a paragraph of plot is not setting up a point you are about to make, cut it or shorten it.

The exception is a review that exists to warn — where a specific plot problem is the point (a mystery whose solution does not track; a non-fiction book whose central case is unsupported). Then the plot detail is evidence, not padding, and it belongs in the "what doesn't work" section where you can explain why.

The spoiler rule

The rule that keeps a review useful to someone who has not read the book:

• You may say a twist exists — "there is a reveal around the two-thirds mark that reframes the first act" — without saying what it is.
• You may not give it away in a review meant to help a reader decide. Write the review a reader can safely read before they buy.
• You may discuss spoilers if you clearly warn first and put the detail below the warning, or after a line like "spoilers below — stop here if you plan to read it."
• Endings are spoilers too. "The last chapter changes how you read everything before it" is fair; "the narrator turns out to be her own mother" is the spoiler, not the review.

Telling a reader a twist lands is a judgement about the book. Telling them the twist is a plot summary that ruins the experience — which is the opposite of what a review is for.

A worked sample review

Here is the same six-part shape, filled in with an invented book. Each part is labelled so you can see the structure at work.

[Hook] Some books you finish and put down; this one I finished and then sat with for an hour, annoyed at how well it had been done. [What it is] The Lighthouse Year (an invented title, used here as a model) is a literary memoir of about 250 pages. [What it's about] A keeper's daughter returns to the island where she grew up to close the family home, and the book is the season she spends sorting through it. [What works] The prose stays dry and specific: the objects in each room carry the memories, so the writing never has to announce its feelings, and the last third earns its quiet rather than declaring its own meaning. [What doesn't] The middle drags in places; a few chapters repeat a similar lesson, and the timeline sometimes jumps without a clear reason. [Verdict] For readers who like patient, understated memoir, this is worth it for the closing chapters alone. Recommended, with patience for the middle.

Notice: the plot is short, the opinion carries the weight, every claim about the writing comes with a reason, and you finish knowing who the book is for.

The same skeleton in three formats

The six parts do not change — only how much of each you keep.

A blog post (600–1,000 words). All six parts, with the evidence expanded. This is the fullest version and the one where the "what works / what doesn't" sections do the most work.

A Goodreads or StoryGraph entry (100–250 words). Hook, one line of what it's about, one strength, one weakness, verdict. Keep the plot to a sentence and lead with your reaction.

An Instagram caption (50–120 words). Hook, one-line premise, the single thing you loved or didn't, and the verdict. If the image is the cover, the caption does not need to describe the book — it needs to tell the reader what you thought.

Five things to avoid

1. Retelling the plot. The most common failure. See the one-quarter test above.
2. Reviewing the author, not the book. Their other work, their persona or their opinions are context, not the subject.
3. A rating with no reason. Stars without a sentence of "why" tell the reader nothing they can use.
4. Spoiling the turn. A twist described is a twist ruined.
5. "It was good." An adjective is not a judgement. Name what worked, name what didn't, and say who it's for.

Frequently asked questions

What are the five parts of a book review?

Most guides list five to six. Ours runs to six: what the book is · what it's about (spoiler-free) · what works · what doesn't · who it's for · the verdict — with a short hook opening the piece.

How long should a book review be?

As long as it needs to make its case and no longer. A Goodreads review can be 150 words; a blog review often runs 600–1,000. If it is long only because you are retelling the plot, cut the plot first.

Do I have to give a star rating?

No. A rating is a shorthand for a verdict you have already written. If you rate, make the sentence next to it explain the number.

Can I write a review without finishing the book?

You can say you did not finish, and that is a legitimate and useful thing to review — but say so plainly and review what you read, not what you assume the rest does.

How do I write a review without spoilers?

Describe effects, not contents: "the reveal reframes the first act" rather than what the reveal is. Warn clearly before any detail you cannot avoid.

What's the difference between a book review and a book report?

A book report proves you read the book and tells a teacher what happens. A book review tells other readers whether the book is worth their time, and why. The report centres the plot; the review centres your judgement.

Where this goes next

Learning to say why a book works — with a line, a scene or a structural choice as the evidence — is the same skill an editor brings to a manuscript. If you are writing a book of your own, our editing and development service is where that eye gets pointed at your draft.

Related guides

Keep going: how to write a book, the types of book editing, book editing services and proofreading services. For a dictionary definition, see book review (Cambridge Dictionary).

Disclosure: Harmony Publishing is a self-publishing company and links to its own services where they match a step. The language guidance here is practical, not a rule for every style guide.