Do Goodreads Reviews Even Matter for Sales?
Eat, Pray, Love, but for early modern Orthodox schismatic sects
They were starovery—Old Believers, the “Russian Amish,” who had fled to the countryside after the Devil made a shambling puppet of the Russian Orthodox Church. Patriarch Nikon of Moscow insisted that he was merely bringing the church in line with the original practices of the wider Orthodox world. The changes were momentous: the sign of the cross should be made with three fingers rather than two, hallelujas should be cried thrice instead of twice, and some prepositions in the liturgy would be changed. To facilitate this reform, it was decreed that the old, outdated prayerbooks should be returned and destroyed, to ensure the unity of Orthodox practice everywhere. But the starovery could not be fooled: the government was coming to take away their prayer books, telling them that the faith of their fathers and grandfathers was a lie. They rebelled; they lost; they fled into the countryside, forming their own communities, separate from the world, keeping the old, true rites.
During the reign of an especially cruel and vicious tsar, one family in Tyumen fled farther than most: deep into Siberia, to the province of Khakassia, near the Mongolian border. Karp Lykov’s brother was murdered in a confrontation with a government soldier, and he knew his family would be targeted next. They loaded up a cart and disappeared into the taiga, hundreds of versty distant from any human settlement, where they could live and pray in peace.
The Lykovs built a proper homestead in the Khakassian taiga, building a home with wood and moss, raising wheat and potatoes in the hardy Siberian soil. In addition to the two young children that came with them, Akulina Lykova bore two more children in the forest, and taught all of them to read and write using their tiny library of prayer books and scripture, practicing their letters with sharpened birch sticks. Everybody had their tasks: Agafia kept track of the calendar over the years and decades, calculating holy days; Dmitry became the family hunter, chasing deer and elk until they collapsed from exhaustion; Savin appointed himself prayer-leader. On nights, they sat around their little campfire and gave thanks to God that they could still live a pure, faithful life in the proper Russian way. True, their life was hard—one winter, mother Akulina starved to death, giving her portions to her children—but they were free to live and worship as God intended, far from the tsar and his agents.
Then the helicopters came.
The peculiar story of the Lykov family and their discovery by Soviet geologists in 1978 has been told in many books and articles—my favorite is Mike Dash’s article for Smithsonian, which he expanded into a book—but has never been adapted into fiction. This was set to change sometime next year with the publication of The Snow Forest by Elizabeth Gilbert (of Eat, Pray, Love fame).
Then something funny happened: The Snow Forest, still eight months away from publication, became one of the worst-reviewed books in the history of Goodreads. Gilbert herself was bombarded with negative messages and emails. The problem, predictably, was that Gilbert was publishing a novel about Russians during Russia’s ongoing invasion of Ukraine. To write about any kind of Russian subject at this time was offensive—a crime on par with playing a Beethoven sonata in 1941, performing a play of Racine’s in 1807, or throat-singing your favorite khöömei in 1215.
So Gilbert wrote, as quoted in the NYT:
“I have received an enormous, massive outpouring of reactions and responses from my Ukrainian readers,” Gilbert said in a video posted on Instagram, “expressing anger, sorrow, disappointment and pain about the fact that I would choose to release a book into the world right now — any book, no matter what the subject of it is — that is set in Russia.”
She continued: “It is not the time for this book to be published. And I do not want to add any harm to a group of people who have already experienced and who are continuing to experience grievous and extreme harm.”
This was a few weeks ago. As of now, The Snow Forest has been placed on an indefinite hold, with no plans to publish it in the future. Like the Lykovs, the book will be waiting in the wilderness, away from the outside world.
There’s not much I can say here on free-speech grounds that hasn’t already been covered ad nauseam in books journalism. The long, complex arc of Russia and its culture is more than a prelude to Putin’s catastrophic war, and most adults can see that. Even if I didn’t have my complicated attachment to Russian culture, I could have told you that you don’t win any prizes by capitulating to angry internet people, even if some of them really are Ukrainians suffering from a brutal, terrifying invasion. Ukraine needs HIMARS rockets, not adjustments to American publishing schedules.
But the thing that baffles me the most here is that this all stems from Goodreads.
The thing that drove Gilbert (and/or her publishers) into such a tizzy in the first place was a swarm of angry reviews on an enshittified, book-themed clone of circa-2009 Facebook. Much like Facebook, Goodreads seemed like a huge deal ten years ago but is now a rapidly-emptying wasteland of trolls, bots, ghost accounts, scammers, and confused old people. Much like Facebook, I still have an account, though I don’t know why and I never use it, except to occasionally see what I was up to in 2013.
Gilbert’s self-imposed book ban has rightly drawn a lot of attention to Goodreads and its shitty moderation of book reviews. The crisis, after all, was kicked off by the fact that the website doesn’t prevent users from posting reviews of books that are not yet commercially available, or sometimes even finished. (George R.R. Martin’s last two Game of Thrones sequels, still many decades from completion, both have thousands of reviews.)
But lost in all the coverage is a much more important question: how important is Goodreads for book sales, anyway? The question is maddeningly difficult to answer.
To be clear, I’m just talking about Goodreads here, not other social media platforms and especially not BookTok. If the revolt against my book had spread to the hordes of sobbing BookTok/Bookstagram influencers, I’d be worried. TikTok has the golden touch for book sales (see here, here, and here), which is why you get “As Seen On BookTok” tables at major bookstores now. What you don’t get, though, are Goodreads-centered tables, despite Goodreads being far more entrenched in the literary world.
Tuesday’s big NYT article on Goodreads reviews and publishing was full of examples of authors who were harassed or shamed on Goodreads, but never actually gets around to how much all this bad publicity actually affects sales. Instead, there’s mostly talk of feelings. And I don’t deny that’s it not fun to have internet strangers piling on to call you a racist, or that this can have an outsized effect on struggling indie authors, but surely Big 5 publishers and their armies of quants have more important metrics than feelings behind their Goodreads strategies. But this is surprisingly hard to figure out.
Back in 2021, I covered a wave of Goodreads scams targeting indie authors, trying to figure out if it was worth paying ransoms to keep your reviews in the positive. Do Goodreads reviews matter? My answer back then was probably, though it was hard to say. The problem with Goodreads, as I covered, is that its data is potentially very valuable for calculating shifts in the book market, which is why Amazon bought it in the first place. The tech world moves fast, though, and in the two years since then, my view on the effectiveness of Goodread have scaled downward dramatically.
Most of what I could find in books journalism on this question was, to put it mildly, useless. This is probably because Goodreads doesn’t publish all of its data, as mentioned above, and more importantly, that most authors hate talking about their terrible, no-good book sales and engagement.
Thankfully, the author Mark Lawrence doesn’t. Every few years since 2015, Lawrence has updated this blog post to track the correlation between Goodreads reviews and sales for a handful of books from the same genre (epic fantasy) with similar publishing backgrounds released around a similar time. In 2015, within this limited dataset, Lawrence found that you could reliably estimate the sales of a book by multiplying its number of reviews by 7.7. That is, a book in Lawrence’s specific niche of modern epic fantasy with 1,000 Goodreads reviews would have around 7,700 total sales. Lawrence revisited this dataset in 2018 and again in 2021, finding that the multiplier has gone down to slightly below 4. This makes intuitive sense: if more readers are on Goodreads leaving more reviews, then the predictive power of an individual review for sales is lessened.
Lawrence, to his credit, doesn’t mistake this correlation for causation, though. His data is a snapshot of sales and reviews, but he remains agnostic on whether one drives the other. It may very well be that sales drive reviews, as more people buying a book would lead to more people having a copy on hand to read and then review. Probably, it’s a complex interaction of both. More importantly, Lawrence’s charts only track the raw number of reviews, not their average rating on Goodreads’s 5-star scale. On his scale, a review is a review, good or bad. In that light, Gilbert’s team might have chosen to view their review-bombing as a positive: engagement is engagement!
Another blogger, English professor Brandon Kempner, also tried to look at this correlation in 2015, checking Goodreads and Amazon reviews against BookScan numbers—the closest thing to a reliable point-of-sale tracking service in publishing, though one that hides its data behind an expensive paywall. Since Publisher’s Weekly often shares BookScan numbers for bestsellers, though, Kempner was able to track a few popular books’ sales against their online reviews. The results were inconclusive, with the sales:reviews ratio stretching from 2.2 sales per Goodreads review to 28.6. Nearly a quarter of Station Eleven’s buyers left reviews, but only 5% of Stephen King’s did. These are useful numbers, though they don’t tell us anything we couldn’t already guess: older authors tend to have older readers who post fewer online reviews; young authors with younger readers have more online reviews. And once again, these don’t really tell us if the polarity of a review matters.
BookScan isn’t the only source I just can’t afford. One 2021 abstract in Information Systems Research did find that “review comments formed after incomplete consumption adversely affect subsequent sales,” but I was unable to secure a copy of the full article and find out how influential these reviews were, and in what ways. More importantly, they didn’t specify where these reviews were coming from—Amazon, Goodreads, or somewhere else?
And other than that, all I could readily find was a cottage industry of self-publishing gurus who say that you NEED (!!!) to be on Goodreads’s ARC program PRONTO!! I’m going to spare you the details. They weren’t helpful.
So we’re not much farther from where we started. Goodreads reviews might help some authors sometimes, or they might be completely pointless, or maybe it’s all about engagement, baby. Nobody is making the case, out loud and public, why reviews are good for your book sales, even though this is presumably very important to Goodreads and publishers. Nobody really knows why some books sell and most don’t. Nobody knows why some books explode on TikTok, or whether become internet rage-bait for a day hurts your sales or boosts them.
To be honest, looking at book sales of any kind for any reason for too long is enough to make you want to run away and live in the Khakassian taiga. Maybe somebody will publish a book about that some day.
From the Archives:
Bonus Track:
Another cover for this week: “Olson,” by Boards of Canada. Video compression did frightful things to the lighting here, but I’m proud of the sound design.
Happy reading!