Madeline Miller's The Song of Achilles was published in 2011 and won the Orange Prize the following year. It had a devoted readership and, in commercial terms, had settled into the long quiet life of a well-reviewed literary novel. Then it surfaced on TikTok, and one books-data site estimated it had sold around two million copies by mid-2022, with most of that arriving a decade after publication.
That story gets told as though a single crying video did it, which is almost certainly too neat — sales figures from third-party trackers are estimates, and a surge like that involves publishers reprinting, retailers reordering and thousands of people posting rather than one person. But it is the thing everybody making book videos is implicitly chasing, and it is worth putting at the front precisely so it can be set aside. Nothing in this article will make that happen to a book you post about. What it can do is make your videos better and keep you out of the avoidable mistakes.
What the format actually asks for
A book video is not a review. A review argues a judgement. A sixty-second book video transmits a feeling and a reason to trust you, in that order.
The structural constraint is that the first two seconds decide everything. Somebody is scrolling, and a video that opens with "hi everyone, so today I wanted to talk about" has already lost. The opening line has to be the point: the feeling, the situation, or the specific thing that makes this book unlike the last forty.
What works after that, in my observation as someone who watches a lot of these without making them, is specificity. Not "this book destroyed me" — everyone says that, and it carries no information — but the particular mechanism by which it destroyed you. The chapter that recontextualises everything. The fact that you knew the ending from the first page and it did not help. One concrete detail that could not be said about any other book.
And then stop. The most common failure is the video that keeps going after it has made its point, usually to add caveats. A recommendation does not need caveats. It needs a reason and an ending.
Spoilers, which are a trust issue
The community norm is stricter than in print criticism, and breaking it is one of the few things that reliably loses an audience.
The practical version: say at the start whether the video is spoiler-free, and mean it. Treat the existence of a twist as itself a spoiler, because telling someone there is a twist changes how they read every page. If you want to discuss the ending, make it a separate video, label it in the first frame and in the caption, and accept that fewer people will watch it.
This costs you engagement in the short term and is the single clearest signal that you are making videos for readers rather than for reach.
The trope vocabulary is a tool, not a gimmick
Outsiders tend to mock the shorthand — enemies to lovers, found family, grumpy and sunshine, dark academia, romantasy — as a reduction of books to labels. Used well, it is nothing of the sort. It is a matching system that does in four words what a genre classification cannot do at all, because it describes the shape of the pleasure rather than the subject matter.
The requirement is accuracy. Tagging a book with a trope it does not deliver is the fastest way to generate a wave of annoyed comments from people who bought it, and the annoyance is deserved. If a romance is billed as slow burn and the couple gets together in chapter four, you have mis-sold something people spent money on.
The other thing worth doing is pairing the tag with the deviation. "Enemies to lovers, except the enmity is actually about something and does not evaporate" is a far better recommendation than the tag alone, because it tells a reader what they are getting beyond the category.
Negative reviews, and where the line is
You are allowed to dislike books publicly. It is a large part of what makes a recommendation account trustworthy, since an account that loves everything is indistinguishable from an advertisement.
Two distinctions are worth holding. The first is between the book and the author. Criticising the writing is criticism. Speculating about the author's character, tagging them into a video about how much you hated their work, or encouraging an audience to go and tell them, is something else, and it does real damage to real people with a speed that short video makes very hard to stop.
The second is about proportion. A debut from a small press and a heavily marketed bestseller are not in the same position, and a large account taking aim at a book with no promotional machinery behind it is not punching in a neutral direction. This is a judgement call rather than a rule, but it is worth actually making the judgement rather than treating all books as equivalent targets.
Disclosure is not optional
If a publisher sent you the book free, if you were paid, if the link earns you a commission, say so — clearly, in the video itself as well as the caption.
This is not merely etiquette. Advertising regulators in several major markets, including the United States, require disclosure of a material connection between an endorser and a seller, and the obligation sits with the creator, not only with the brand. Beyond the legal position, it is the thing that protects the value of your recommendations. An audience that discovers retrospectively that your favourite books of the year were all gifted will discount everything you say afterwards.
Review copies themselves are ordinary and nothing to be embarrassed about. They arrive through the standard digital galley platforms, through publisher publicity lists, and increasingly through direct approaches once an account has any size. Accepting one carries no obligation to post, no obligation to be positive, and no obligation to post on a schedule, whatever an over-eager publicist implies. Say that up front and the relationship stays clean.
About the algorithm
Most advice in this area is confident about a ranking system that is proprietary, undocumented and changed frequently. Posting times, magic hashtag counts, the claim that some format is favoured this quarter — treat all of it as folklore. Nobody outside the company knows, and the people selling courses about it know least of all.
What is structurally true, and does not require guessing: watch-through matters, so short and tight beats long and loose. Spoken words and on-screen text are searchable, so saying the title and author aloud and putting them on screen makes a video findable months later. Captions make a video usable with sound off. And a back catalogue that is consistent in subject is easier for a new viewer to evaluate than a feed that is books one week and something else the next.
Everything beyond that is inference dressed as method.
The part that goes wrong quietly
Making content about reading changes reading. The pressure to have something new to say pushes towards fast, short, high-emotion books and away from the slow ones; the pressure to post pushes towards finishing things you would otherwise abandon; and at some point a number of people find they are reading in order to have read.
There is no clever solution. What helps is keeping some reading entirely off the account — a book nobody will hear about, chosen for no reason except that you want to — and being relaxed about posting less. Accounts that disappear usually do so because their owner burned out, not because the algorithm turned on them.
If you are an author
The most common mistake is trying to pass as a reader account and being found out. The second most common is treating every video as a sales pitch.
What appears to work is showing the process, having actual opinions about other people's books, and being a person rather than a marketing function. Whether it is worth your time depends entirely on whether you enjoy it, because the returns are unpredictable and the effort is constant. Writers who dislike video should spend the time writing.
One structural note
Everything above is platform-independent, which matters, because a following that exists only inside one app is held on terms you do not control. If any of this becomes important to you, put something durable alongside it — a newsletter, a list, anywhere you can reach people directly. Formats change. Recommending books to people who trust you does not.
Tags
Taresh Sharan
support@sharaninitiatives.com