Wednesday, November 11, 2026 · The BookTok Intelligence Daily
Inside the Industry

Molly Stern on the $200M Imprint Arms Race Nobody Is Talking About

The Zando founder on poaching editors from Flatiron, why the Big 5 advance bubble has a shelf life, and what BookTok actually demands from a publisher in 2026.

By · · 9 min read
Molly Stern on the $200M Imprint Arms Race Nobody Is Talking About

Molly Stern has spent the better part of three decades making publishing uncomfortable for the people who run it. She built Crown into a cultural force, then walked away to found Zando in 2020 with a model that still makes legacy houses nervous: lean staff, fat advances where it counts, and a direct line to the creator economy that most imprint directors are still trying to map on a whiteboard. When we met at the Crosby Street Hotel on a Tuesday morning in early July, she had just returned from Bologna and was about to fly to Seoul. The foreign rights machine, she said, has never moved faster — or gotten messier. We talked for ninety minutes. This is the edited version.

You’ve watched three major editor reshuffles happen in the first half of 2026 alone. What is actually driving the talent exodus from the Big 5?

The honest answer is two things happening simultaneously. The first is structural: the consolidation hangover from the failed PRH-S&S merger never really resolved itself. You had a generation of senior editors at Penguin Random House who were told, essentially, wait your turn for five years, and when the merger collapsed they looked around and realized the queue hadn’t moved. Those people are restless. Some of them are brilliant. And boutique houses — Zando, Flatiron, Celadon, and, frankly, the new independents being funded by private equity — are calling them directly.

The second thing is BookTok. I know that sounds reductive, but hear me out. A senior editor at a Big 5 house is working inside a marketing infrastructure that was built for a pre-creator world. They acquire a book they love, they fight for budget, they get assigned a publicist who is handling fourteen other titles that month, and then they watch a twenty-three-year-old in Austin with 800,000 followers single-handedly move 40,000 units of a competing title in a week. That is a deeply demoralizing experience if your institution doesn’t have the agility to respond. We do. That matters to editors who want their books to actually reach people.

“The Big 5 marketing machine was engineered for a world where you controlled the chokepoints — the review pages, the retailer co-op, the morning shows. Those chokepoints don’t exist anymore. BookTok doesn’t care about your Kirkus starred review.”

Zando closed what sources are calling a nine-figure backing round earlier this year. What does that capital actually buy you in the current acquisition environment?

I’m not going to confirm a specific number because we’re a private company and I have lawyers who get paid to keep me from doing that. What I will say is that the capital environment for serious independent publishers right now is genuinely extraordinary, because sophisticated investors have figured out something the legacy houses learned too late: IP is the asset class of the decade, and books are the cheapest IP on the market.

Think about what a seven-figure advance for a debut novelist actually buys you. You are acquiring film rights, TV rights, audio rights, translation rights, merchandise rights, potential game rights — the entire stack — for what Netflix spends catering a single episode of a prestige drama. The math is absurd in favor of the publisher who moves quickly. So yes, our capital position lets us write checks that would have seemed reckless in 2019. I don’t think they’re reckless. I think they’re correctly priced for the moment.

What it also buys is speed. We preempted two manuscripts in Q2 that went to auction at other houses first. One of them — I can’t say which one yet — had four Big 5 imprints in the room. We won because we could move in forty-eight hours with a full offer including a BookTok marketing budget line item that was larger than what most houses spend on their entire social strategy for a season.

“I preempted a manuscript in May that had PRH, HarperCollins, and Macmillan in the room. We won on speed and a BookTok budget that made their social line items look embarrassing. The author’s agent called it ‘the fastest clean close she’d seen in twenty years.'”

Let’s talk about the ghostwriting question. There are persistent rumors that several of the biggest BookTok-native launches of the past eighteen months involved significant uncredited collaboration. How serious is this problem?

It’s real, and the industry is not handling it honestly. I want to be careful here because ghostwriting itself is a completely legitimate and ancient practice — celebrity memoirs, business books, a significant portion of the thriller market has always worked this way. Nobody hides the ball when a former president publishes a memoir. The problem is specifically in the BookTok-creator-to-author pipeline, where a creator with 2 million followers gets a seven-figure deal based on audience capture and then the manuscript is delivered by someone the contract calls a “developmental collaborator” but who wrote the entire book.

I’ve been in rooms where this was discussed openly as a strategy. “She doesn’t write, but she has the audience, and we’ll support her through the process” — which is sometimes a euphemism for “we’ll hire someone to write it and she’ll do the BookTok.” That’s not inherently wrong if it’s disclosed. The problem is the disclosure almost never happens, and readers on BookTok feel genuinely deceived when it comes out. And it always comes out. The Reddit threads are ruthless.

What I find more troubling, honestly, is the AI question layered on top of this. We’re now at a point where some of those “developmental collaborators” are themselves using AI drafting tools to produce first passes. So you have a creator who didn’t write the book, working with a collaborator who partially didn’t write the book, and somewhere in that chain is a language model that nobody is disclosing. That is a legitimacy crisis waiting to happen at scale. One major expose — and someone is going to write it, probably this year — could crater audience trust in the entire creator-to-author category.

You’ve been outspoken about AI in acquisitions. Where exactly does Zando draw the line?

We use AI aggressively on the business side. Comps analysis, rights tracking, translation cost modeling, marketing performance prediction — all of it. I would be a fool not to. Our data team has built a signal dashboard that ingests BookTok velocity metrics, Goodreads shelf movement, and Substack engagement simultaneously, and it’s genuinely useful for deciding when to move on an acquisition and at what price. That’s AI as infrastructure. I’m completely comfortable with it.

What we will not do is acquire a manuscript that was substantially written by a generative AI system. We have a clause in our author agreements now. And I’ll be honest, we’ve passed on projects in the past six months where the writing felt statistically smooth in a way that raised flags. One of those projects sold elsewhere for a significant advance. I hope I’m wrong about it. I don’t think I am.

The deeper issue is that BookTok readers are actually quite good at detecting inauthenticity. Not AI specifically — they’re not running detection software — but the emotional register of AI-assisted prose tends to be even. There are no rough edges. And BookTok, at its core, is an obsession economy built on rough edges. The books that explode on that platform are the ones that feel like they were written by someone who was slightly too much. AI doesn’t do “slightly too much.” It does competent. Competent doesn’t break the algorithm.

“BookTok is an obsession economy. It runs on rough edges, on the feeling that a book was written by someone who was slightly too much. AI does competent. Competent doesn’t break the algorithm.”

The foreign rights market has gone through a significant restructuring since AI translation tools matured. What does that mean for how you’re selling rights now?

Bologna this year was a different fair than it was three years ago, and I say that having attended every year since the late nineties. The structural change is that AI translation has collapsed the time-to-market gap between English publication and foreign editions from eighteen months to, in some cases, eight weeks. That sounds like good news for everyone and it is, mostly — but it has also completely repriced translation rights deals at the lower end of the market.

Small language markets that used to pay modest but reliable advances are now offering less, because the cost of producing the book has dropped and they know it. Turkish, Dutch, Polish, Czech — those markets have repriced downward by something like twenty to thirty percent in the advance range, depending on the category. Romance has been hit hardest because the translation quality for commercial romance is genuinely very good now and publishers in those markets have less incentive to bid aggressively. They’ll just wait for a smaller deal.

What has gone in the other direction is the premium market. Korean, Japanese, German, French, Brazilian Portuguese — those deals are larger than I have ever seen them, because the cultural prestige signal of a BookTok-native phenomenon is now globally legible in a way it wasn’t before. Korean publishers in particular have become extraordinarily sophisticated about reading TikTok velocity data as a leading indicator. I had a Korean editor at Bologna quote me our own Goodreads numbers back to me. She knew the data better than half my team.

Last question. The advance bubble — you’ve watched it inflate in real time. Is there a correction coming, and who gets hurt when it lands?

There is always a correction coming. The question is what triggers it and who is overexposed. My honest read is that the houses most at risk are the ones who have been paying eight- and nine-figure advances for creator-to-author projects that are essentially audience arbitrage: you’re not buying a book, you’re buying a follower count and hoping it converts. Some of those bets will pay off spectacularly. A meaningful number will not, and the write-downs will be significant.

The specific vulnerability I would watch is the multi-book deal signed at peak creator heat. A creator signs a three-book deal in Q4 2024 at the top of their algorithm moment. By the time Book Two is due, their follower engagement has moved, their audience has aged into different reading habits, and the BookTok algorithm has promoted six new voices into their lane. The first book performs adequately. The second book underperforms. The third book is quietly killed. The house has eaten a loss, the creator’s literary career is effectively over, and an agent somewhere is having a very difficult conversation. I have seen this pattern play out four times in the past two years. It’s going to play out more.

What I think survives the correction — and this is genuinely what we’re building at Zando — is the model where you’re acquiring actual manuscripts that can stand independent of their author’s social footprint, and then building a BookTok campaign around the book’s own emotional architecture rather than the creator’s face. The book has to be the thing. The platform is distribution. When publishers forget that order of operations, they get hurt. Some of them are about to be reminded.

“The multi-book creator deal signed at peak follower heat is the ticking exposure nobody on the earnings call is discussing. I have watched it blow up quietly four times already. It is going to be loud the next time.”

Sponsored Intelligence
How forward publishers run book ads in 2026: AdGPT — built for the BookTok economy.
See Inside

More in Inside the Industry

View All →