Finishing a book and wanting to tell somebody about it is one of the better feelings going. The trouble starts when that enthusiasm meets a blank page. Most first drafts turn into a retelling with a verdict bolted on the end: here is what happened, here is whether I liked it. That is a book report. A review tells a stranger whether the book deserves their evening. The gap between the two is craft, and craft can be learnt.

A review is an argument, not a receipt.

Retelling the plot instead of reviewing the book

The plot summary is scaffolding, not the building. Readers can get the setup from the blurb in ten seconds; what they cannot get anywhere else is your judgement. New reviewers often spend two-thirds of a review walking through the story in order, then bolt three lines of opinion onto the end.

Try this test. Cover your summary and read only what is left. Is there an argument? If it collapses into “I liked it”, the summary is carrying too much of the review.

Two or three sentences is usually plenty: the situation, the pressure on the main character, and the turn that makes the book interesting. Then stop. Resist explaining how it resolves. A reader who knows the ending has less reason to buy the book, and a reader who has already finished it does not need reminding.

Watch for the other version of this mistake, too: the review that recounts the plot and the themes, so the reader gets the argument of the book without any of the experience. A review should tell someone what it felt like to read the thing, not only what it said.

What counts as a spoiler

A spoiler is not only the ending. It is anything the author deliberately held back. Ask whether a reader would mind learning a detail before page fifty. If the answer is yes, leave it out or flag it plainly: “skip the next paragraph if you want to go in cold.”

This cuts against the instinct to quote your favourite lines. A great line from page 200 is often great precisely because of everything that built up in front of it. Quote the opening instead — first lines are written to be seen early.

Praise that says nothing

“Beautifully written.” “The characters felt real.” “I could not put it down.” These are the sounds of a reviewer being polite rather than useful, and they fill a great many early reviews.

Vague praise usually has one cause: it is hard to explain why something worked. The fix is to name the mechanism. Instead of calling the prose beautiful, point to what makes it so — sentences that stop short and leave a gap, or dialogue that never quite answers the question it was asked. Instead of saying the characters felt real, describe a single moment that made you believe in them.

• Vague: “The protagonist was really relatable.”
• Specific: “Mira rehearses arguments in the shower and loses every one. It is an embarrassing little detail, and it made me trust her.”

The same applies to criticism. “I could not connect with it” says nothing about the book. “The middle third makes the same argument in three different kitchens” says a great deal.

One more rule of thumb: if a sentence could be pasted onto any other book without changing a word, delete it. “A compelling read” survives that test. “The chapters are short enough to tempt you into one more at midnight, and I gave in four nights running” does not.

Writing for yourself instead of the reader

Every review has at least two readers: the person deciding whether to buy the book, and the person who has already read it and wants company. A third — the author — is watching quietly from the wings.

Decide who you are writing for before you start. A genre blog can assume familiarity with the conventions. A general audience cannot, and will not thank you for jargon or references to the author's back catalogue.

The book's intended audience matters too. Judging a cosy crime novel by the standards of literary fiction is not clever; it is a category error. Say who the book is for, then say whether it delivers to that reader. That is the question most people actually have.

The same mistake shows up in the small print. If you received a review copy, say so. If you know the author, or once shared a table with her at a conference, say so. Readers are not naive about free books, but they do object to discovering the connection later. One honest line costs nothing and buys a great deal of trust.

Reviewing the book you wanted rather than the one you got

This is the mistake keen readers make most often, and the hardest to spot in yourself. You open a novel expecting the author's usual style and get something else. You wanted a thriller that moves; you got a slow, interior book about grief with one body in it. You wanted the sequel to be more of the first book; the author wanted to write something new. Now every line of the review is really a complaint about your expectations, not about the writing.

The tell is the phrase “I wanted more…” Finish that sentence honestly. More of what? More plot, more romance, more of the last book? Sometimes the answer is a fair criticism — a thin middle act is a thin middle act, whoever wrote it. But often the answer is simply that the book is not the book you had in your head, and that is not the author's fault.

Two questions help. Did the book do what it set out to do? And did it do it well? Praise or blame it on those terms and keep your own unmet wishes out of the verdict. If the mismatch is worth mentioning — and sometimes it is, especially for readers of an author's earlier work — frame it as information rather than accusation: “Readers of Salt Road should know that this one is quieter.”

The verdict problem

New reviewers tend to assume a review has to end in a score. It does not, but if you are giving one, make it match the words above it. Five stars for a book you spent three paragraphs dismantling is not generous; it is confusing. Readers notice the mismatch and start trusting the whole review a little less.

If a book is genuinely mixed, say so. “Excellent first half, flabby ending” is a useful review. So is “not for me, but exactly right for someone else” — provided you explain who that someone is. A review that refuses to be useful is not subtle. It is just unfinished.

A short checklist before you post

1. Does the review say something a reader could not get from the blurb?
2. Can you point to one specific sentence, scene or decision as evidence for your main claim?
3. Is there a spoiler you have not flagged?
4. Have you said who the book is for?
5. Would you still write it this way if a friend had written the book? If not, soften the tone and keep the substance.

None of this takes long. The first few reviews are slow, because you are learning to notice your own reactions while you are still having them. After a while it becomes automatic: you finish a book and you already know, roughly, what your argument is and which two scenes you will use to make it. That is when reviewing stops feeling like homework and starts feeling like the conversation you wanted in the first place.

Photo: Pexels / Pixabay