Nobody at ScrollWorks expected it. We had printed 2,000 copies, which felt optimistic at the time. The book was a quiet literary novel about a woman who restores old photographs for a living, set in a small town in Vermont that the author had invented from scratch. The cover was nice but not flashy. The author had no social media following to speak of. Our marketing budget for the title was, to put it charitably, modest.

Six weeks after publication, we were out of stock. By the eight-week mark, we had rushed two additional print runs and could not keep up with orders. By the end of the year, that quiet little novel had outsold everything else on our list combined. I am still not entirely sure how it happened, and that uncertainty is part of what makes the unexpected bestseller one of the most fascinating phenomena in publishing.

I want to write about this honestly, which means admitting upfront that publishers are terrible at predicting which books will break out. We have decades of data, industry expertise, sophisticated pre-publication review processes, and relationships with booksellers who have excellent instincts. And despite all of that, the books that become surprise hits are, by definition, surprises. If we could predict them, they would not be unexpected.

The publishing industry has a dirty secret that nobody likes to talk about at conferences. Most of our revenue projections for individual titles are wrong. Not slightly wrong. Sometimes spectacularly wrong, in both directions. We overestimate books that we think will be big, and we underestimate books that turn out to be the ones readers actually wanted. The industry keeps functioning because the hits, when they come, are profitable enough to cover the losses from the misses. But the process of figuring out which will be which? That is mostly guesswork dressed up in spreadsheets.

I have been thinking about this a lot lately because we have had two unexpected successes in the past three years, and looking at them side by side has taught me some things I did not expect to learn.

The first thing I have learned is that unexpected bestsellers almost never come from nowhere. When I say our Vermont novel was a surprise, I do not mean there were zero early signals. There were signals. We just did not know how to read them. The book had received a couple of very enthusiastic early reviews from independent bookstores that participated in our advance reader program. One bookstore in Portland, Oregon, hand-sold over 200 copies in the first month. The owner told me later that she had been pressing the book into customers’ hands with a level of personal conviction she reserved for maybe three or four books a year.

That kind of bookseller passion is incredibly powerful, and it is almost impossible to manufacture. You cannot pay for it. You cannot engineer it through a clever marketing campaign. It happens when a particular book connects with a particular reader in a way that makes them feel genuinely compelled to share it. Word of mouth is the oldest and most effective marketing channel in existence, and it runs entirely on authentic enthusiasm.

The second signal we missed was in our own correspondence. The author had been receiving emails from readers at a rate that was unusual for a debut novelist with a small press. Within the first two weeks, she had gotten over fifty personal emails from readers, not just “I liked your book” messages, but long, detailed letters about how specific passages had affected them. One woman wrote three pages about how the novel’s depiction of grief had helped her process the loss of her mother. A retired photographer wrote to say the book had captured something about his profession that he had never been able to articulate himself.

The author mentioned these emails to her editor in passing, almost as an afterthought. Nobody at ScrollWorks connected the dots at the time. In retrospect, fifty heartfelt reader emails in two weeks for a debut novel from a small press is an extraordinary signal. It suggests the book is hitting an emotional frequency that resonates deeply with the people who find it. And deeply resonant books get recommended.

Here is what I think happens with an unexpected bestseller, at least based on what I have observed. The book finds its first few hundred readers through normal channels: bookstore placement, a handful of reviews, the author’s personal network. Among those first few hundred, a disproportionate number have an intense reaction. They do not just enjoy the book; they feel compelled to tell someone about it. So they recommend it to friends, family, book clubs, coworkers. Some of them post about it online. A few of those posts gain traction.

At some point, the book crosses a threshold where the recommendations are generating more new readers than the publisher’s own marketing efforts. This is the tipping point, and when it happens, the publisher’s role shifts from driving demand to simply trying to keep up with it. Our job becomes making sure the book is in stock, that bookstores can reorder quickly, and that we are not leaving readers unable to buy a book they have already decided they want.

We failed at this during our first unexpected hit. We ran out of stock for almost two weeks while we waited for the second print run. During those two weeks, the book was functionally unavailable at many retailers. I am convinced we lost sales permanently during that gap. Some percentage of people who wanted to buy the book during those two weeks moved on to something else and never came back. The momentum of an unexpected bestseller is fragile, and a stock-out can break it.

Since then, we have adjusted our approach. We now have a standing agreement with our printer that allows us to trigger a reprint within 72 hours for any title that is selling above a certain threshold. This costs us a little more per unit because we are not batching our print orders as efficiently. But the insurance against a stock-out is worth it. Lost sales during a breakout moment are the most expensive kind of lost sales, because each lost sale also represents a lost recommendation.

The second lesson from our unexpected bestsellers is about the role of timing. Both of our surprise hits were published during periods when their subject matter happened to align with something the culture was collectively processing. The Vermont novel came out during a period when a lot of people were thinking about small-town life, solitude, and the question of what makes a community. It was not written in response to those cultural currents. The author had started it years before. But it arrived at a moment when readers were primed for exactly what it offered.

I do not think you can plan for this kind of timing. Publishing a book takes so long, typically 12 to 18 months from acquisition to publication, that trying to chase cultural trends is a fool’s errand. By the time your trend-chasing book hits shelves, the trend has usually moved on. What you can do is publish books that are genuinely good, that deal with themes and emotions that are always relevant to human experience, and hope that the timing works out. Sometimes it does. Most of the time, it does not.

The third thing I have learned is how differently the publishing ecosystem treats a book once it starts showing unexpected strength. Before our Vermont novel broke out, getting review coverage was like pulling teeth. We sent review copies to dozens of publications and received mostly silence. After the book started selling, those same publications came to us. Reviewers who had ignored the advance copy suddenly wanted to write about it. Media outlets that had passed on author interviews were now requesting them.

This is frustrating, and I think it reveals something honest about how literary media works. There is a strong herd instinct. Once a book is perceived as successful, it becomes much easier to get coverage of it, which in turn makes it more successful. The rich get richer. Meanwhile, the books that might have been equally good but did not get that initial spark of organic enthusiasm remain invisible.

I do not say this bitterly. It is just the reality of how attention works in a market where thousands of new books are published every week. Reviewers and media outlets have limited space and have to make choices about what to cover. A book with demonstrated reader enthusiasm is a safer bet for coverage than an unknown quantity. I get the logic. But it does mean that the path to becoming an unexpected bestseller is front-loaded with difficulty. You have to generate that initial enthusiasm with very little institutional support, and then once you have it, the institutional support suddenly materializes.

Let me talk about what an unexpected bestseller does to a small press internally, because this is something that does not get discussed much. When a book dramatically outperforms expectations at a publisher the size of ScrollWorks, it creates both opportunities and pressures that can be destabilizing if you are not prepared for them.

On the opportunity side, a bestseller generates cash flow that can fund future acquisitions. It raises the profile of the press, making it easier to attract good manuscripts. It gives you leverage with distributors and bookstores. Authors who might not have considered a small press suddenly see you as viable. All of this is good.

On the pressure side, there is a temptation to try to replicate the success. After our Vermont novel hit, we received a flood of manuscripts that were essentially “quiet literary novel set in a small New England town.” Authors and agents had noticed what was selling for us and were trying to give us more of the same. And honestly, we were tempted. The data said this type of book could sell. Why not publish more of them?

We resisted that temptation, and I am glad we did. Because the thing about an unexpected bestseller is that its success was not really about its category or setting. It was about the specific quality of that specific book meeting that specific moment. Trying to reverse-engineer a formula from one data point is bad statistics and bad publishing. The next book that breaks out for us might be a historical epic. It might be a collection of essays. It might be something we cannot even categorize. The only consistent predictor I have found is quality combined with authentic emotional resonance, and you cannot systematize either of those things.

There is also the question of what an unexpected bestseller does to the author’s career. This is complicated territory. On one hand, a breakout book opens doors that were previously locked. The author gets invited to festivals, receives foreign rights inquiries, and suddenly has a readership eagerly awaiting their next book. On the other hand, that eager readership creates pressure. The second book has to live up to the first, and that expectation can be paralyzing.

I have watched this play out with two authors now, and the pattern is similar. There is an initial period of elation and disbelief. Then a period of anxiety about the next book. Then a long stretch of hard work during which the author has to somehow write something genuine and personal while knowing that tens of thousands of people are waiting to judge it. My advice to both authors was the same: ignore the sales numbers, ignore the expectations, and write the book you need to write. Whether they managed to follow that advice is their story to tell.

One more thing about unexpected bestsellers that I think is worth mentioning. They change your relationship with risk. Before our first surprise hit, I was fairly conservative about acquisitions. I wanted to see clear market comps, identifiable audiences, and realistic sales projections before I would sign a book. After watching a book with none of those things become our biggest seller, I loosened up.

I now give more weight to my own gut reaction when reading a manuscript. If a book makes me feel something genuine, if it lingers in my mind after I finish it, if I find myself wanting to describe it to people at dinner, those are the signals I trust most. They are not scientific. They are not data-driven. But they are exactly the signals that predicted our unexpected bestsellers, and they are exactly the signals that our spreadsheets and market analysis completely missed.

The publishing industry spends a lot of time and money trying to figure out what readers want. We analyze trends, study bestseller lists, commission market research, and attend conferences where experts share their predictions for the next big thing. All of that has some value. But the unexpected bestseller is a humbling reminder that readers are individuals, not data points, and that the books that move them most deeply are often the ones that nobody saw coming.

Our latest titles, from Still Waters to Echoes of Iron, were each acquired because they made someone on our team feel something real. Whether any of them becomes the next unexpected hit, I honestly have no idea. And I have come to think that not knowing is part of what makes this work worth doing. Every book we publish is a bet on the possibility that it will find its readers. Some bets pay off in ways we never imagined. That possibility, even when it remains unrealized, is what keeps me showing up.