A recommendation is a small promise: someone is telling you that the hours you spend with this book will be worth something. Most of the time that promise is honest. Sometimes it is a marketing decision wearing an enthusiastic hat, and the difference is usually visible before you pay — if you know where to look. None of the checks below take more than a few minutes. Together they will save you a shelf of half-read disappointments.

1. Who is recommending it, and what's in it for them?

Start by naming the source. A publisher's jacket copy is sales material, written to make a book sound unmissable rather than to describe it accurately. Praise from another novelist on the cover is usually a favour, occasionally a genuine rave, and rarely a detailed assessment. A friend who reads the way you read is worth more than fifty anonymous five-star reviews.

Then ask the sharper question: has this person actually finished the book? Booksellers who have read it, reviewers with a byline and a word count, and readers who can name a character are all giving you something concrete. Endorsements, listicles and "readers also bought" suggestions tell you about marketing, not quality.

2. Does the blurb say what the book actually is?

A good blurb answers three things quickly: who the book follows, what they want, and what stands in the way. If you finish reading it and still cannot say what happens in the first fifty pages, the copy is hiding something — often that the book is thinner than its pitch.

Watch for the language of mood rather than substance. Lines like this are a warning rather than an invitation:

A sweeping, unforgettable story of love, loss and the secrets we keep — perfect for fans of everything you have ever enjoyed.

That could describe several thousand novels. Useful blurbs mention specifics: a setting, a voice, a structure, an argument. For non-fiction, ask whether the book promises a claim you can test or merely a subject you already find interesting. A history of a period you like can still be a poor history of it.

Check the practical details jackets tend to play down, too. Is it the fourth in a series? Is it a translation, and by whom? Is the "new" edition actually a reissue? If the description reads like a book you would not pick up, believe it.

3. Are the reviews describing the book, or just applauding it?

"Loved it!" tells you about the reader. "Loved it, though the middle dragged and the twist felt unearned" tells you about the book. Hunt for reviews that mention structure, pacing, a named character, or the way the author handles a difficult idea. If a review could be pasted onto any book in the genre without anyone noticing, it is not doing any work for you.

A few habits that pay off:

• Read the middling reviews first. Two and three stars are usually written by people who finished the book and have something precise to say.
• Look for repeated complaints. One reader calling a novel slow is taste; six readers calling it slow is information.
• Check the dates and the spread. A wall of praise posted in the same week as publication deserves a raised eyebrow.
• Ignore the score, weigh the detail. A thoughtful three-star review beats a gushing five-star one nearly every time.

4. What did that award actually reward?

A prize sticker means a book was noticed by a particular jury in a particular year. That is worth something, but it is narrower than "good". Prizes reward debuts, or a specific genre, or a certain kind of literary novel, or a book that happened to catch the mood of its moment.

Find out what the award is for, and if you can, read the citation explaining why it won. Shortlists are often more useful than winners, because they show the range of books the judges thought belonged together. A prize for a first novel is not a promise that the second will be as strong.

5. Does this critic's taste line up with yours?

Build a small list of reviewers whose opinions have matched your own experience. Two or three is enough. The test is not whether they liked the books you liked, but whether their reasons make sense to you and whether they flagged the same flaws you noticed.

A reviewer who disliked a book you loved, and explained why clearly, is still useful: you now know how they read and can calibrate accordingly. Pay attention to their comparisons. If a critic says a novel is for readers of a writer you have tried and disliked, treat that as a warning.

6. What do the opening pages tell you?

Read the sample before you buy, or read the first few pages standing in the shop. Ten or fifteen is usually enough. Read for voice rather than plot: do you enjoy the sentences, do you trust the narrator, do you want to know what happens on page eleven? A recommendation can put a book in your hands, but only the prose keeps it there.

Then apply the patience test. Ask whether you would still want this book in three months, once the launch coverage has moved on and nobody is talking about it. If you would pick it up from a table with no sticker on the cover, you have found something worth buying. If not, you were probably buying the conversation rather than the book.

Before You Click Buy

Five quick questions, and how to use the answers:

1. Can you say in one sentence what the book is about, in your own words rather than the jacket's?
2. Did at least one review mention something specific that appeals to you?
3. Did the sample make you want page eleven?
4. Has the person recommending it earned your trust before?
5. Would you still want it if nobody mentioned it again?

Four or five yeses: buy it and enjoy it. Two or three: add it to a list and revisit it next month, when the noise has settled. Serious hesitation is rarely wrong — it is your own reading taste telling you something the marketing has been careful to avoid. Trust it, and your shelves will stop filling up with books you meant to like.

Photo: TheUjulala / Pixabay