Last week, on the eve of the Frankfurt Book Fair, the largest and oldest in the world (it dates back to the fifteenth century), AmazonCrossing, Amazon Publishing’s translation imprint, announced it would be committing $10 million over the next five years to publishing more works in translation. “We launched AmazonCrossing five years ago to introduce readers to voices of the world through English-language translations of foreign-language books,” AmazonCrossing’s publisher Sarah Jane Gunter said in a statement. “While we are now one of the largest publishers of translated literature in the United States, translated fiction is still a tiny fraction of new publications.”

Over the past five years, only Dalkey Archive, the uber-literary small press that has published books by authors like Carlos Fuentes, Viktor Shklovsky, and Danilo Kiš, has published more works in translation than AmazonCrossing. This year, AmazonCrossing plans to publish “77 titles from 15 countries and 12 languages” in the United States, which will almost certainly dwarf the output of Dalkey and its ilk. And, with this new $10 million commitment, the number of works in translation published by AmazonCrossing should continue to soar. Which means that AmazonCrossing will almost certainly be the largest publisher of translated literature in the United States for at least the next five years.

Of course, $10 million over five years for works in translation is not a world-shaking announcement—after all, Simon & Schuster just gave comedian Amy Schumer close to that amount for just one book. And, while AmazonCrossing did announce a few interesting tweaks to its operations (more on those in a moment), it is mostly committing to continuing to do what works, just on a slightly grander scale. Still, though $10 million over five years will not turn AmazonCrossing into a publishing powerhouse, it still has important implications for translators and for readers.

Despite Amazon’s contentious relationship with the larger publishing industry (Jeff Bezos, the corporation’s founder, once instructed his employees to “approach… small publishers the way a cheetah would pursue a sickly gazelle” in negotiations), the company runs a number of highly successful small and medium-sized publishing houses, including Thomas & Mercer, which focuses on thrillers, Little A, which publishes more literary fare, and AmazonCrossing.

Amazon’s imprints are somewhat isolated, both geographically and commercially, from the wider publishing world. Based in Seattle, not New York (where all major—and most minor—publishers are based), Amazon Publishing typically distributes its books only through Amazon.com. As a result, you won’t see them on the New York Times bestseller list, which discounts books that are only available from a single retailer, or on the subway: most are sold in electronic form at price points between $1.99 and $4.99.

It wasn’t necessarily supposed to be this way. In 2011, Amazon attempted to break into New York publishing, hiring industry veteran Larry Kirshbaum in 2011 to lead "Amazon Publishing's East Coast Group" and launch New Harvest, a somewhat sinister sounding imprint that was intended to compete with traditional literary houses. Kirshbaum's efforts alarmed the publishing industry: many feared that the company, having cornered the market on selling books would corner the market on producing them as well .Kirshbaum and Bezos, for their part, both insisted they had only modest ambitions: Bezos told Businessweek that the company had no intention “to become Random House or HarperCollins,” though he admitted that “I think people have a hard time believing that.”

No one in New York believed Bezos. As Brad Stone writes in The Everything Store, “a kind of industrywide immune response kicked in” soon after Amazon announced its expanded publishing arm. Kirshbaum had been given a directive to land big fish and the resources to do so, but he struggled to sign well-known authors to Amazon Publishing, in part because signing with one of its imprints effectively locked you into one retail platform: most retailers, including Barnes & Noble, don’t usually stock books published by Amazon in retaliation for the retailer’s many predatory practices. Foregoing 35% of the market was a risk most agents and authors did not care to take, and while Kirshbaum eventually landed a string of reasonably well-known nonfiction authors, including Tim Ferriss and Penny Marshall, the resuts were mixed at best. Ferriss’s book sold well, but didn't meet expections and undoubtedly took a substantial hit after Barnes & Noble refused to carry it. (Ferriss did himself no favors when he claimed this was censorship.) Marshall was given an $800,000 advance for a memoir that is now regarded as one of the biggest flops of the 21st century.

Kirshbaum left Amazon in 2013, and the retailer’s publishing arm began a slow retreat to Seattle shortly thereafter. When Ed Park—the New York-based novelist and editor who launched the Little A literary fiction imprint—left the company last fall, it was tempting see his departure as a nail in Amazon Publishing’s coffin: the New York Times, for instance, ran a headline that described the move as a “major setback” for Amazon publishing. Park himself admitted that he would not “miss [the] obstacles” he faced with authors, agents, and retailers as an Amazon editor.

Amazon Publishing tried to play the publishing industry on its own turf and lost. Few prominent authors were wrested away from Big 5 publishers, few books won awards or even received significant attention (books published by Amazon are rarely reviewed at all in traditional outlets), and, though a few books sold well, sales weren’t enough to convince agents that publishing with Amazon was worth the risk. While Amazon has continued to disrupt how books are sold and distributed over the past five years, Amazon Publishing bears little of the responsibility for those changes.

But it would be a mistake to judge Amazon Publishing by its failure to make a splash as a more traditional, New York-based publisher—or by its inability to attain the kind of cultural relevance those publishers covet. As Amazon’s efforts to establish more traditional, literary-minded, New York-based publishers failed, its Seattle-based imprints which were focused on the Kindle and more commercial fiction were showing another path forward. In 2012, after Kirshbaum had floundered in New York for a year or so, Amazon’s imprints began doing things a little differently. Per a Seattle Weekly feature about Amazon Publishing that appeared last year:

Not only did Amazon ask them to change the type of authors they were looking for, but the company insisted upon a different way of evaluating them. Going with your gut was out. Data was in. What kind of track record does the author have? The category of books? What type of people bought them?

The former staffer describes it as a “consumer-brand approach that would not be unfamiliar at P&G [Procter & Gamble],” and also as a “West Coast” one. (Wait, isn’t the West Coast supposed to be laid-back and freewheeling, not corporate and data-obsessed? Such is the power of Amazon that it’s giving the entire coast a different hue.)

In some ways, AmazonCrossing was the opposite of the more conventional house Kirshbaum attempted to build. That house would do what many publishing houses do: pay authors large advances for books that would (hopefully) sell a substantial numbers of copies. While it wasn’t quite intended to mirror a traditional publishing house, it was intended to resemble one; at the very least, it was supposed to compete for the same books.

AmazonCrossing, which preceded Kirshbaum by a year, resulted from a different strategy: identify niches that are being underserved by mainstream publishers. In fact, it’s fair to say that over the last five years, AmazonCrossing and a number of its fellow imprints have created a kind of third way between traditional publishing and self-publishing. The curation one expects from a traditional house is there, as is the sense of quality control. The economics, however, closely resemble those of self-publishing: risk is largely assumed by the author, who receives a substantial cut of an insubstantial retail price. Of course, that risk is somewhat mitigated by the assumption that the marketing arm of the only platform a book published by Amazon is expected to sell on, the Kindle, will heavily support the book in question.

Indeed, the entire system is built around the Kindle. Amazon’s substantial (some would say near-monopolistic—the company commands at least 65% of the ebook market) market share has blocked its access to big name authors and crucial markets and prevented it from running a more traditional publishing house, but that same size has allowed it to create a publishing ecosystem in its own image. Amazon is fond of lecturing publishing houses about the various ways that they’re outdated and inefficient, and its own publishing houses offer a glimpse of what Amazon believes the larger publishing landscape should look like.


Overproduction has been a fact of life in the publishing industry for decades, probably centuries, but nothing historically can compete to the flood of content, particularly self-published content, that followed the advent of the Kindle. Many inside the industry complain that there are “too many books,” but Amazon has always taken an opportunistic attitude to this situation. As Open Letter Books’ Chad W. Post told me, “their line on it has always been that they exist to make as many books as possible in as many forms as possible.” Looking for these niches was part of that wider philosophy. The first niche Amazon Publishing identified was commercial, translated fiction.

American publishers are averse to publishing works in translation, which are seen as being too costly: you not only have to pay a foreign publisher for the rights to publish it and a translator to translate it, but it probably won’t sell very well. As a result, America has a famously poor record publishing works in translation. It’s said that only 3% of books published in the United States are works in translation, though even that number may be too high. 587 works in translation were published in the United States last year, according to figures compiled by Post. That number is up from 340 in 2010, but it’s doubtful that it accounts for 3% of all books published in the U.S. (The numbers may work if you’re only counting works of fiction, but even then, you have to exclude works self-published on Amazon and elsewhere.)

That doesn’t mean that literary fiction in translation doesn’t sell. One can point to countless examples from the past 15 years: the Japanese novelist Haruki Murakami is wildly popular, the Chilean novelist and poet Roberto Bolaño has posthumously sold hundreds of thousands of copies in America, and Elena Ferrante and Karl Ove Knausgaard, an Italian and Norwegian, respectively, have become literary phenomenons recently. All four of those authors, admittedly, were popular elsewhere before they were popular in America.

That last point is key. When asked why translated fiction doesn’t sell, many in the publishing industry argue that translated fiction is too challenging or too intimidating to American audiences. I think, though, that this may be getting things backwards: a great deal of translated fiction struggles to sell in America not because it’s translated, but because it’s challenging in any language. For every Roberto Bolaño, there are a number of intelligent South American novelists whose work didn't make a big splash in their home countries—and didn't resonate in the US.

And because translation has a reputation for selling poorly, the publishing houses that translate the most tend to be smaller. The biggest publishers in translated fiction in America that aren’t AmazonCrossing—Dalkey Archive, Seagull Books, and Europa Editions—are all independent, small presses. Of the ten biggest publishers of translated literature in America, only two are imprints of conglomerate publishers: FSG, which has a long tradition of publishing works in translation, and Atria, which mostly publishes books by celebrities, and they finish 8th and 9th, respectively.

In other words, the current publishing landscape is not geared not towards the most commercial foreign writers, but the most interesting ones. (Which is not to say that there aren’t plenty of exceptions; indeed, many foreign-language books with solid commercial prospects—not to mention English-language ones!—also fail to find a readership.)

AmazonCrossing has a literary seasoning—Post expressed particular admiration for the Ukrainian novelist Oksana Zabuzhko, who has published three novels with AmazonCrossing—but the imprint is fundamentally different from nearly all of the other leading publishers of translated works in America. A New York Times story that appeared earlier this year, “Who Is the Biggest Publisher of Foreign Literature in the U.S.?,” begins to explain why.

AmazonCrossing’s biggest hit so far is probably The Hangman’s Daughter, written in German by Oliver Pötzsch. Its success illustrates how AmazonCrossing capitalizes on its parent company’s global footprint to decide which books to translate. This historical mystery, published by Ullstein Verlag, sold unevenly in German bookstores but managed to gain traction with readers on Amazon’s German site.

“We noticed that Oliver’s books were getting rave reviews from customers, who said things like ‘I can’t put it down,’ ‘It’s incredible,’ ‘I want to read more,”’ said Sarah Jane Gunter, the publisher of AmazonCrossing.

The Hangman’s Daughter has sold over a million copies (mostly electronically) in English since it was first published in 2010. The editors at AmazonCrossing weren’t out to identify books that were getting rave reviews, or works by writers who are winning international awards. Nor were they necessarily looking for international blockbusters, which are often expensive to acquire. Instead, they were looking for books that readers were connecting with, which they could figure out using Amazon data.

Editing and acquiring unpublished works in your own language is a cerebral experience. “If you’re going to try to do American books like a normal American publisher,” Post told me, “the process of deciding what to publish is very much in someone’s mind. You have to figure out ‘Is this a good book, will this find an audience.’ That’s a skill that’s mental and not exactly quantifiable.” Data can help identify the kinds of books that are selling, but not the books themselves—making a decision about what to publish requires that an editor bet on a book’s sales potential. AmazonCrossing’s approach, though, can ease that discomfort. Think of it as a publishing version of “Moneyball”—except Amazon is Billy Beane and The Hangman’s Daughter is Scott Hatteberg.

Even without exact numbers, there is more than enough evidence to suggest that AmazonCrossing has found a level of success. It’s published over 600 books over the last five years, including the million-selling Hangman’s Daughter; and Amazon doesn’t often make lengthy, multi-million dollar commitments like the one it announced last week to projects that aren’t bringing in revenue. AmazonCrossing has found this success by avoiding competition with traditional publishers for authors and books—instead, it seeks out stories that connect with readers in places other publishers aren’t looking, effectively using American publishers’ abysmal translation rate against them. And, because American publishers translate so little, foreign agents and authors are more willing to work with Amazon than their American colleagues—they have fewer options.

But just as American agents who worked with New Harvest had to balance financial concerns with the question of access, foreign agents and authors who work with AmazonCrossing know that they’re getting locked in. Like many—if not all—of Amazon’s imprints, AmazonCrossing is focused almost exclusively on the Kindle, where the vast majority of its sales occur. This has its advantages: Amazon’s imprints rarely if ever have to worry about the kinds of things that occupy the time of a traditional publishing houses: everything from standard publicity efforts to author events and major reviews are off the table. These books have one job, so to speak: to reach the right Kindle readers.

In terms of fiction, Amazon’s imprints have had far more luck with genre titles than with literary fiction, and AmazonCrossing is no different—the successes it most often touts are mysteries, thrillers, and romances. Genre, after all, is king on the Kindle—successful self-published books also tend to be genre, rather than literary. Which makes sense: Amazon has always advocated an aggressive brand of price elasticity: it has consistently argued that the volume that follows lower price points will lead to more revenue than higher, more conservative price points, and Amazon’s imprints have tended to target niches with price-conscious but voracious readers. Books published by Amazon’s imprints tend to more closely resemble those favored by self-publishers (typically between $1.99-$4.99) than those preferred by publishing houses (typically between $9.99-$14.99).

These low prices are another one of Amazon Publishing’s peculiarities. Regardless of one’s affinities, publishing sales are almost exclusively analyzed in terms of units sold, not revenue, despite the fact that revenue is what really counts. If one assumes that The Hangman’s Daughter sold a million-and-a-half copies and that these copies were exclusively ebooks, AmazonCrossing earned between $3-$7.5 million, of which it would then owe the author a significant portion—at least half. These are terms that Amazon can offer only because it’s operating small- and medium-sized houses and printing a negligible number of copies of books.

These are prices that are designed to move Kindle copies, not mitigate risk, and they better. While information about AmazonCrossing’s terms is hard to come by, literary agent Jane Dystel told the Seattle Weekly last year that “We’re seeing advances somewhere in the range of $10,000 or $20,000,” from Amazon’s non-translation imprints. Instead of large advances, authors are offered significantly higher royalty rates. Royalties on books published by traditional houses typically max out at 15% of list price for hardcovers, 10% of list price for paperbacks, and 25% of net receipts for ebooks. Royalty rates given to Amazon Publishing authors differ, but they tend to be far closer to the 70% royalty given to those who self-publish on Amazon than those offered by traditional publishers. (Translators who work with Amazon are also getting different terms: based on conversations I’ve had with people in the know, Amazon pays translators a competitive rate and that AmazonCrossing pays all its translators a royalty, which is pretty uncommon.)

What kind of translated fiction is selling is up for debate, however. In an interview with the New York Times, Post argued that AmazonCrossing fulfils an important cultural role: “To fully understand a culture, to understand literature the whole way, you need more than just the best books. You need the book about Spanish history along with the book about zombies.” Judging a culture only by its most esoteric output gives you a distorted impression of that culture; there’s more to Argentina than Aira. But I do think that it could be argued that this move is less culturally significant than it would have been a few decades ago. Cultures are more connected and more stories are being shared across geographical borders. The significance of AmazonCrossing may not be that it allows us to have a unique insight into Iceland or Mozambique or Indonesia, but that it’s benefiting from the internationalization of culture. As people around the world consume more and more of the same media, narratives and styles may become less regional and more international, widening the reach of a publisher focused on international books with wide appeal. 

AmazonCrossing recently announced a new project in mechanization: a “portal” through which translators, foreign publishers, and foreign authors can submit work to be considered for publication by AmazonCrossing. The portal itself seems to be a bet that the international market they’ve tapped into is substantial—that there are hidden gems awaiting translation. But anyone who’s tackled a slush pile for a publishing house or publication knows that nearly everything that’s submitted is garbage-—AmazonCrossing may hope that it’s working in a specialized enough field to provide a modicum of quality control. They may see it as a way to get bargains from desperate authors, translators, and publishers. They may merely see the benefit in communicating a certain degree of openness—after all, Amazon’s imprints are gatekeepers that can’t be seen as gatekeepers.

And what about that $10 million? Where exactly is it going? In a press release, AmazonCrossing announced that the money would go “toward fees paid to translators over the next five years and increasing the countries and languages represented on the AmazonCrossing list.” When I reached out for comment that somewhat cloudy statement was reiterated, though with an indication that AmazonCrossing also intends to increase the number of titles it publishes, something it’s done every year since its inception. The money for translators will presumably pay more translators, given the increase in production, but as Susan Bernofsky notes on her indispensible blog Translationista, there’s no reason for translators to expect a sudden windfall as a result of this announcement—rates are just as likely to decrease, she argues.

Given the wiggle room and the timing—it was clearly engineered to garner attention at Frankfurt, where Amazon has struggled to get meetings before—it’s entirely possible that the announcement is significantly less significant than it appears. For all we know, AmazonCrossing committed a similar amount of money to acquiring and translating international literature last year. $10 million, after all, is a rounding error for Amazon. What does seem to be significant, however, is that AmazonCrossing is proving that Amazon can run a publishing house its way—based on data analysis and staggeringly low prices—and succeed.