About three years ago, I received an email from a reader named Margaret. She was 74 years old, lived in rural Vermont, and had just finished reading one of our titles. Her email was four paragraphs long, carefully written, and contained a sentence that I’ve thought about nearly every day since: “I don’t think you realize what this book did for me, and I’m writing because I think you should.”

That email changed how I think about publishing. Not because it was flattering (though it was), but because it made me confront a blind spot I didn’t know I had. I had been thinking about readers as an abstraction, a market segment, a demographic. Margaret wasn’t a demographic. She was a person in Vermont with arthritis in her hands and a reading light that she had to replace every six months because she read so much at night. She had specific opinions about font size, paper quality, and whether a book’s spine would survive being passed between friends. She had been reading for seventy years and had extremely clear ideas about what made a book worth her time.

I wrote back. We started corresponding. Over the next year, Margaret sent me notes on three of our books, and each note contained observations that our entire editorial team found valuable. She noticed things we’d missed. She articulated reactions that our focus groups had never surfaced. She was, without trying to be, one of the best readers I’d ever encountered.

Margaret was the first, but she wasn’t the last. Over the past three years, reader feedback has fundamentally reshaped how ScrollWorks operates, and I want to tell that story because I think it matters for anyone who cares about the relationship between publishers and the people they publish for.

The feedback we weren’t getting

For our first four years, our primary feedback channels were the ones every publisher uses: sales data, reviews in professional publications, and the occasional reader email or social media mention. These are useful, but they have significant blind spots.

Sales data tells you whether people bought the book. It tells you nothing about whether they finished it, what they thought about it, or whether they’d buy the next one. A book can sell well on the strength of its cover and marketing and still be a disappointment to the people who actually read it. Conversely, a book can sell modestly and be adored by every person who reads it. Sales data can’t distinguish between these two outcomes.

Professional reviews tell you what professional reviewers think, which is valuable but narrow. Reviewers are not representative of the general reading public. They read hundreds of books a year, which means they approach each one with a set of expectations and references that most readers don’t share. A reviewer might praise a book for its “formal innovation,” while an actual reader in Michigan just wants to know if the characters feel real and if the ending is satisfying.

Social media is noisy and self-selecting. The people who post about books online skew younger, more urban, and more interested in performative literary culture than the average reader. BookTok and Bookstagram are useful for certain kinds of books, but they represent a small fraction of the reading public, and their aesthetic preferences don’t always align with what works on the page.

What we were missing was direct, thoughtful, sustained feedback from actual readers. Not influencers, not reviewers, not the algorithmically amplified voices. Regular people who bought our books, read them, and had opinions about them.

Starting the reader panel

After my experience with Margaret, I started thinking about how to create a more structured way of gathering reader feedback. The result was what we now call the ScrollWorks Reader Panel, which is essentially a group of about sixty readers who receive advance copies of our books and provide detailed feedback before and after publication.

The panel isn’t a focus group in the traditional sense. We don’t give people surveys with Likert scales and multiple-choice questions. Instead, we send them the book, ask them to read it at their own pace, and then invite them to write to us about whatever struck them. No prompts, no structure, no minimum word count. Just: read this, and tell us what you think.

The responses we get are extraordinary. They range from single-paragraph emails to ten-page letters. Some readers focus on the prose. Others focus on the characters. Some talk about how the book made them feel. Others talk about how it connects to their own experiences. A retired teacher in Ohio wrote three pages about how one of our novels reminded her of a student she’d taught thirty years ago. A truck driver in Texas sent a voice memo (not an email, a voice memo) describing the scene that made him pull over at a rest stop because he couldn’t see the road through his tears.

This kind of feedback is not available through any other channel. It’s not quantifiable, it’s not scalable, and it’s not efficient. It is, however, real. And real feedback from real readers has changed our publishing decisions in ways that I want to describe specifically, because I think the specifics are what matter.

How reader feedback changed our editorial process

The first major change came with font size. I know this sounds mundane, but it’s a good example of how reader feedback can surface issues that publishers are blind to. Several panel members, independently, mentioned that our books were printed in a font size that strained their eyes. These weren’t elderly readers with failing vision (well, some were, but not all). They were readers in their 40s and 50s who did a lot of their reading at night and found our 10-point type exhausting after thirty minutes.

We looked at our books with fresh eyes and realized they were right. Our standard body text was 10 on 13 (10-point type with 13-point leading), which is on the small side for trade fiction. We’d been using that spec because it reduced page counts, which reduced printing costs. The savings were real but small: maybe 30 to 40 pages per book, translating to about 15 cents per copy in production cost.

We switched to 11 on 14.5. The books got slightly longer and slightly more expensive to print. They also became significantly more comfortable to read, especially for sustained reading sessions. Multiple panel members noticed the change immediately and thanked us for it. One wrote: “I finished the whole book in two sittings instead of my usual five. My eyes didn’t get tired.” That single change probably did more for reader satisfaction than any editorial decision we’ve ever made.

The second major change involved pacing. Multiple readers told us, in various ways, that the middle sections of our literary novels tended to “sag.” They used different words for it: “slow,” “heavy,” “where I put the book down for a few days.” This was consistent enough across different books and different readers that we couldn’t dismiss it as individual preference.

We took this feedback to our developmental editors and asked them to pay specific attention to mid-book pacing in future projects. Not to make our books faster, necessarily, but to ensure that the middle sections were doing active work, progressing the emotional or thematic argument rather than merely filling space between the beginning and the end. This became a formal part of our developmental editing checklist, and I believe it’s improved our books noticeably.

When Catherine Voss was revising The Last Archive, we shared (with her permission) some of the pacing feedback we’d received on earlier books. She restructured her third act in response, moving a revelation earlier and adding a subplot that maintained tension through what had been a quiet section. The finished book is better for it, and that improvement traces directly back to readers who cared enough to tell us what wasn’t working.

The cover conversation

Book covers are traditionally designed without much reader input. The publisher, the author, and the designer collaborate, sometimes with input from the sales team, and the result is whatever emerges from that conversation. Readers see the cover for the first time when the book is published. By then, it’s too late to change anything.

We started showing cover concepts to our reader panel about two years ago, and the results have been illuminating. Readers see covers differently than publishing professionals do. We tend to evaluate covers aesthetically: Is it well-designed? Does it communicate the genre? Is it distinctive on a shelf? Readers evaluate covers emotionally: Does it make me want to pick this up? Does it feel like a book I’d enjoy? Would I be comfortable reading this in public?

That last question, “Would I be comfortable reading this in public,” was one I’d never considered before a panel member raised it. She explained that she read on the subway during her commute and was self-conscious about the covers of her books. A cover that was too obviously “literary” made her feel pretentious. A cover that was too commercial made her feel embarrassed. She wanted covers that were beautiful but not trying too hard, which is actually an excellent design brief.

We’ve adjusted our cover process to incorporate reader feedback at the concept stage. This doesn’t mean readers make the final decision, because design is a professional skill and I trust our designers’ expertise. But reader input has eliminated some covers that looked great on a computer screen but would have been problematic in the real world, and it has confirmed other covers that our team was uncertain about.

What readers taught us about marketing

One of the most valuable things our reader panel has taught us is how people actually discover books. The publishing industry spends an enormous amount of time and money trying to answer this question through market research and data analysis. Our panel gave us the answer for free: people discover books through other people.

Not algorithms. Not social media ads. Not email newsletters. People. Specifically, people they trust. A friend who says “you have to read this.” A bookseller who puts a book in their hands. A family member who gives it as a gift. A book club member who champions it at a meeting.

This finding, which we’ve heard from our panel members so consistently that I consider it essentially proven, reshaped our marketing strategy. We shifted resources away from digital advertising (which was expensive and produced mediocre results) and toward activities that put books into the hands of people who would recommend them. More advance reader copies. More bookseller outreach. More book club partnerships. More author events at stores where people could meet the writer, hear them read, and leave feeling personally connected to the book.

We also started what we call “reading ambassador” kits: packages we send to enthusiastic readers (identified through our panel and through organic outreach) that include a finished copy of an upcoming book, a few extra copies to share, and a handwritten note from the editor explaining why we’re excited about the book. The recipients aren’t obligated to do anything. We don’t ask them to post reviews or share on social media. We just give them books and trust that if the book is good enough, they’ll talk about it.

They do. Consistently. Our reading ambassador program is, by a significant margin, the most cost-effective marketing activity we’ve ever done. And it only works because we trust readers to be honest advocates rather than treating them as marketing channels to be optimized.

When reader feedback is wrong

I want to be honest about something: reader feedback is not always right. Or, more precisely, reader feedback is always honest but not always useful. Sometimes a reader’s reaction to a book is deeply personal and doesn’t generalize. Sometimes what a reader wants is not what a reader needs. Sometimes the thing that makes a book uncomfortable or challenging is exactly the thing that makes it important.

We had a panel member who consistently found our literary fiction “too slow.” She wanted more plot, more action, more forward momentum. Her feedback was sincere, and I appreciated it, but acting on it would have meant abandoning the kind of books we exist to publish. She was the wrong reader for our list, and that’s okay. Not every book is for every person.

We also received feedback on Echoes of Iron suggesting that certain historical sections were “too depressing” and should be softened. James Whitfield and I discussed this, and we both felt strongly that softening those sections would be a form of historical dishonesty. The book is about hard things, and representing those hard things truthfully is part of its moral purpose. We kept the sections as written.

The skill, which I’m still developing, is knowing when to listen and when to hold firm. I listen when feedback points to a craft issue, something that could be done better without compromising the book’s vision. I hold firm when feedback asks a book to be something other than what it is. The line between those two responses isn’t always clear, and I get it wrong sometimes. But having the feedback at all, having real readers’ voices in the conversation, makes every decision better-informed.

Margaret’s legacy

Margaret, the reader from Vermont who started all of this, passed away last spring. Her daughter emailed me, which was how I found out. The daughter mentioned that Margaret had told her about our correspondence and said it was one of the things that made her last years richer. She asked if I would send a copy of our next book to her mother’s reading group at the library, because the group intended to keep meeting in Margaret’s memory.

I sent them all of our books. Every title in the catalog. With a letter that I wrote and rewrote about six times before I was satisfied that it said what I needed it to say.

Publishing is a business. I don’t lose sight of that. We have budgets, cash flow targets, distribution agreements, and all the other machinery of commerce. But the reason the machinery exists, the fuel that makes it run, is the connection between a writer’s mind and a reader’s heart. Every decision we make should serve that connection. And the people best positioned to tell us whether we’re succeeding are not critics, not algorithms, not industry analysts. They’re people like Margaret. People who read because reading is how they make sense of being alive.

We owe them more than a good product. We owe them our attention, our respect, and our willingness to change when they show us a better way. That’s what I’ve learned from three years of listening to readers, and it’s what I intend to keep learning for as long as ScrollWorks exists.

Written by the ScrollWorks Media editorial team. Our readers are our best editors, and we’re grateful for every word they share with us.