From BookBub:
At BookBub, we connect books to readers, and readers to books. One of the components that drives this connection is the blurb we write for each book featured in our Featured Deals email. A successful blurb caters to the settings, characters, and tropes our readers love — so to write such a blurb, we need to learn as much as possible about our readers’ tastes.
To do this, we A/B test many of our blurbs, which lets us evaluate the performance of certain words, phrases, punctuation, or other blurb elements. Through this analysis, we can see what our readers engage with — and what turns them away. This post will highlight some recent A/B test results that you may find useful as you write and improve your own book descriptions.
. . . .
We run A/B tests by creating two different versions of a blurb: the A version is the control, and the B version has a slight variation. Most users see the A version, but a randomly selected group gets the B version instead. By comparing each blurb’s click-through rate (CTR, or how likely users were to click on the book), we measure the impact of the change we’re testing.
. . . .
1. Call out authors’ accolades
Readers respond well to mentions of an author’s accolades, including awards the author has won. Blurbs that named prestigious, genre-specific awards boosted CTRs by up to 25%, with an average increase of 5%.
. . . .
2. Avoid including too many characters’ names
It seems intuitive that one great way to help readers connect with a book is to introduce them to characters, calling out the main characters by name. However, test results show that this may not be the case. Especially in books with several main characters, names in the blurb hurt its performance. The example below seems drastic, but the trend is consistent across all of our tests. On average, blurbs with 3 or more names saw a 10% lower CTR than the nameless variations.
. . . .
7. Call out characters’ ages in Chick Lit
If blurbs shouldn’t mention too many characters’ names, what kind of information should they include? The performance of certain character attributes can be hard to predict. Furthermore, readers’ engagement with these attributes often varies by genre — which is why running your own tests with your own readers is so important! Here’s a great example of this: in Chick Lit, mentioning a character’s age consistently helps drive engagement. Including the heroine’s age boosted CTR by an average of 9%in this genre.
Link to the rest at BookBub and thanks to A.R. for the tip.
1. I saw no sample size information, no confidence interval, no statistical foundation for any of their claims. Thus to me, this is clickbait.
2. I won a writing award, and so what? Once I picked up a book at the library that purported to be a novelization of a Texas cattle drive to supply George Washington’s Continental Army with beef. Huh! Interesting subject. Won some award for Westerns. Said so on the jacket. Read the first two pages. My thought — “Is it really this bad?” Read the first five pages. “Yep, it’s really this bad. This clown cannot write.” Awards mean nothing.
3. Like A C Flory, I read a lot of sf. My typical daily read includes one sf short story anthology and one sf novel. I do not consider the Hugos or the Nebulas to be indicators of quality anymore. Not since 2000. Both have become political. Frell ’em.
4. Jada Pinkett Smith threw a hissy fit ’cause hubby Will was not nominated for an Oscar. The perfect response came from Denzel Washington. DW has two Oscars. Doesn’t matter. What matters? Box office. DW is box office. So is Will Smith. Box office. That’s the only award that counts.
I’m a sci-fi devotee, so I’d have to say that mentioning a very prestigious award – like a Hugo or Nebula – would make me stop for a second look, especially if I was not familiar with the author.
By contrast, I found the quotes from other sources, none of which I recognized, quite annoying. Personal taste, I guess.
As a reader, I thought the authors wrote their own blurbs. So when I see punctuation errors, typos, and bad word choices, I avoid the book, thinking those problems are indicative of the writing. Now that I know Book Bub writes the blurbs, I need to change that attitude.
And I agree with Jamie – leave the author’s awards off, please. The only award I have any interest in at all would be an award for that book itself, and I’d still rather just learn about the story.
I agree about leaving the awards out. If I see certain awards, I skip the book because I know from experience that isn’t the type of material I enjoy reading.
All you folks talking about awards or not – YOU are not your customer. The only proof is running tests. If you want more tests, run them, but they are just giving statistical test on readers, not writers, and those are the people who matter. At least, if you want to sell books.
BookBub, like other bookselling sites, uses whatever blurbs the publisher (or author-publisher) provides, I believe (and technically, it’s cover copy or product description, but “blurb” appears to be in common usage).
For many self-pubbed authors (and small publishers), I find that typos, poor grammar, misspellings, etc. ARE indicative that they wrote them themselves and the writing inside the book will be the same. Some authors, however, hire blurb writers in the same way they hire cover artists, in order to help the work look more professional. Articles like this give terrific advise, and it’s something any author can learn. It becomes a question of how much the author’s time is worth, and if this is a job that can be done better, faster, and cheaper by a pro.
I am skeptical of 1 — I hate when blurbs, which are usually about 300 characters, waste precious space saying the author is award winning. I don’t care. I want to know what the story is about.
I notice they only compare “Bram Stoker Award Winning” to “award winning,” which I suspect is only revealing that if you do waste space on awards, you should be specific. Personally, some awards are likely to repulse me as a buyer. I wonder if they test that? Could be revealing …
2 makes sense; there’s no point in naming characters readers aren’t attached to just yet. I suspect this rule is different for an ongoing series: Sherlock Holmes faces off against archnemesis Moriarty.
Now, 8, use “Lord” vs. captain, I think is genre specific. They say that “Jessamine pines for Lord Christopher” goes over better than “Jessamine pines for Christopher,” but they aren’t sure that military titles resonate. They admit they need to test this more, and I agree, because I suspect “Captain” works just fine in science fiction. And possibly also historical fiction or alternate history.
Specifically, I’m wondering if everyone knows the fellow who fought at Trafalgar as “Admiral Nelson” and your blurb begins with “Lieutenant Nelson,” would this keep you from needing to explain “this story is about the young Admiral Nelson”? More testing, please.
Anyone interested in learning more should definitely click over. They give a concrete example for each one, showing how one change affected the clickthrough rate.
I wonder if they did an A/B test on the phony-baloney awards to see if buying yourself one would be effective. (They used the Bram Stoker in their example, which is credible.)