Finding Publishing Opportunities in an Era of AI Disruption: Q&A with CEO Keith Riegert

I first saw Keith Riegert in action at the 2024 US Book Show in New York City, where he gave a compelling presentation on the capabilities of AI that I recapped in this newsletter. But it wasn’t the demo itself that was most remarkable for me. Rather, it was his attitude toward it: reluctant necessity. He didn’t take a lot of joy in showing what his companies already do with AI. Mostly, his talk felt like an apologetic warning to the industry: Ignore AI tools and you may find yourself without a job in the future.

Riegert is associated with a multitude of publishing-related businesses and serves as the CEO of Ulysses Press and Velo Press. I associate him most strongly with Perfect Bound, which began as an online marketplace to match publishers with printers in hopes of creating faster and cheaper print runs. It has since expanded what it offers. Last year, it announced a partnership with Simon & Schuster to offer distribution for independent publishers. 

There’s more: As announced in February 2025, Riegert is now the CEO of the Stable Book Group, a collective of independent publishers that includes Ulysses and Velo. Stable is meant to help independent publishers join forces and become more profitable through sharing of resources. And as you’ll learn in our conversation, this joining of forces is closely tied to Perfect Bound.

The following conversation took place in late March 2025 over Zoom. It has been edited for length and clarity.

---

Jane Friedman: Since February, you’ve established the Stable Book Group and the Publishers Cooperative. You remain CEO of Ulysses Press and Velo Press. You’re also CEO of Perfect Bound and co-founder of Pacific & Court, a publicity firm. You serve on a few boards. So, if I were coming to you without any knowledge of publishing and how these companies have strong connections, I would say that you’re a serial entrepreneur. So I’m wondering, do you consider yourself that?

Keith Riegert: I guess so. I feel like entrepreneurship is a bit of a disease, and if you get it, then it sticks with you. And it is not something that, honestly, I would wish on most people.

Why is that?

It is the hardest you will ever work in your life for almost no money. You know, I have never been paid at Perfect Bound, and that’s just the way that it is when you’re a founder. And a lot of people who think about starting their own business do not realize the enormous sacrifices that they’re about to make, and how scary it is. You are taking on not only the responsibility for yourself and your family, but for everybody that you hire. And, you know, if it’s a publishing company, all of your authors and all of those relationships, that really rests on your shoulders.

Given the number of companies that you’re involved with or that you’ve started, I imagine you have to trust these teams to get things done or be in leadership roles to get things done. Is there a particular place right now that you’re focusing your time and attention?

The best way to start and grow a business is finding the right team and be okay with pushing down as much responsibility as possible. The team that I have at Perfect Bound, primarily Claire Davey and Yiyi Huang, are incredible, and they know the platform better than I do at this point. Having other people be able to shoulder responsibility is critical.

Right now, I am very focused on Perfect Bound and The Stable Group. Ulysses and Velo have incredible teams running them, doing every piece of the business. And I see my role [as] thinking about what’s coming next and strategizing how this is all going to play out over the next couple of years.

That seems like a really loaded phrase—“how this is all going to play out over the next couple years.” Can you tell me more about what you mean by that?

We have very big ambitions for The Stable. This has all been a roadmap that we put out for ourselves several years ago, and now it’s starting to really pick up and come together. I do not consider Perfect Bound and The Stable to be completely operating on their own. Perfect Bound is now becoming the tech that is undergirding The Stable, and all of our companies, and will allow us to operate sales and distribution services within The Stable using Perfect Bound. So, all of these things are definitely coming together. I’m really excited to build a new way of independent publishing. I just think that there is so much opportunity.

The barriers to entry for almost every form of arts and entertainment has collapsed over the last 20 years with websites like Etsy, Spotify, KDP. The industry has not really caught up to those changes. Putting in brand new, cutting edge technology onto very creative business models is going to be enormously fun and I hope successful.

When you say cutting-edge technology, is that a reference to Perfect Bound?

So yeah, Perfect Bound started as just the printing marketplace, and that was something that I just wanted to exist. As an independent publisher, we had such a hard time reaching out and trusting that we could use a brand-new printer for anything. So we were using the same four printers for all of our books. And when the pandemic came along, it was very hard to get a book [to a new printer]. So how do we make it a lot easier for independent publishers to find and work with new printers?

I partnered with the Vimbly Group, which owns and operates 12 different SaaS platforms. … So we started to expand it out where you could take a book from conception to distribution all in one place.

Until a couple of years ago, we were doing our royalties on a spreadsheet that would constantly crash because it was so large. We were tracking our inventory on spreadsheets. We’re hosting our metadata with our distributor. We don’t really own anything and nothing is talking to each other.

I have a background in data analytics, so I am very driven to make everything into a relational database where we can pull as much information as possible about anything that we need. I think that that has been really sorely lacking in the industry. So now with Perfect Bound we’ve got our title management system that Onix feeds out to any distributor or point of purchase that you want. Then we’ve got sales and inventory tracking where we’re fully mapped to take in sales from any distributor or any point of sale. So if you have sales at Shopify and KDP and Lightning Source and Simon & Schuster, all you have to do is take that data, drop it in, and it automatically calculates your total earnings and then flows into the royalty accounting platform, which allows you to automate your royalties. We have the printing marketplace so you can completely understand your margins. And then the final part is PubMatch, which we’re currently rebuilding to offer a global rights solution for independent publishers and their agents.

So that is the foundation that we have built, and what I see over the next couple of years, especially as AI really comes online and we have AI agents, is if you give those agents a scaffolding on which to operate, you are going to put yourself in a position where fewer mistakes [are] made, with a much more efficient operation. Imagine AI agents on Perfect Bound being able to track your inventory, make a request for a print quote based on that inventory level, actually calculate [and send out] all of your royalties on an ongoing basis so your accounting team knows exactly what to accrue, and be able to manage foreign rights sales very efficiently through the platform. Having everything in one place is the way to do it. That is where we’re headed with Perfect Bound.

If I’m a small publisher that’s not yet participating in these services, do I have to buy into all of these systems to take advantage of what you’ve built?

Everything is available a la carte. We are not trying to come in and say, “You need to use every part of this or nothing.” We want to keep it open to everybody regardless of their needs. Especially if you’re an established independent publisher, not only do you already have your systems in place, but there’s a significant cost to switching, and that is not a cost that most independent publishers are going to be ready to make right away. Again, as an independent publisher, we’re building the things that I just wish we had 10 or 15 years ago, and in my mind should have always existed.

Could you talk about the increased costs of print production in comparison to revenue, specifically for the types of independent publishers that use Perfect Bound?

So I actually crunched the numbers for Ulysses a couple of months ago, and our median cost of goods per unit has gone up 75 percent since 2010, and our median list price has gone up 3 percent. That, to me, really summarizes the biggest problem in the industry. The margins have really been squeezed. Printing is expensive. The tariffs and confusion around what is happening with trade in the US is pretty terrifying.

When we started using Perfect Bound, our paper-over-board, full-color prices dropped 18 percent year over year, and our black and white dropped 11 percent. So after 10 years of that line going straight up, in the last year, it started to dip.

I think the real answer is we’ve got to figure out how to increase the perceived value of a book, which I think has been artificially low for decades. And the value is there, you see it. A book is now a luxury item. It is a break from your screen.

I think the best thing that came out of the pandemic, this sudden spike in inflation, was that it sort of broke stagnating prices in printing, and you are seeing $30 hardcovers. Pricing has gone up. There’s still pressure from retailers to keep your prices under a certain point.

I lay a lot of the blame on ebooks for the price stagnation. I did a study when I was in school, looking at how ebook pricing artificially impacted every single format of book over the next 10 years. It started with mass market. Mass market just stopped in its tracks. You couldn’t raise prices anymore. Over the subsequent years, you saw that effect start to move into paperbacks and then hardcovers. The perceived value of a book when it’s a 99-cent ebook or a free ebook does a lot of damage to our ability to compete as a form of entertainment on the higher end, which is one of the reasons I really love the movement toward special editions and limited editions.

How big do you think The Stable Group will get? Do you have this vision for having dozens of publishers under that umbrella or is there a natural limit to how big it can get?

We want to grow. We have a target of $50 million in top-line revenue within the next five years, and most of that is going to come through acquisition. We will probably bring in another 12 publishers over the next couple of years.

Going back to just how difficult it is to be an independent publisher, and how unbelievably labor-intensive this industry is, in my mind, based on the margins and what profit looks like in publishing, the optimal ratio of employees to revenue, to top-line revenue, is one employee per $1 million in revenue. And I don’t know a single independent publisher that is a million-dollar company that has one person doing all of the things that need to be done.

To me, the real success of the Stable is being able to share resources that independent publishers otherwise have to recreate themselves at their own small organization level, to be able to share those resources, whether that’s accounting or production or marketing. Getting a positive ROI on your social media marketing, or even your Amazon advertising, is very difficult if you are not able to optimize based on a very large backlist. We want to market [all publishers’ titles] at the same time, in the same campaigns, and make it work for each publisher and the organization at large.

I want to go back to the employee-to-revenue ratio you expressed. Tell me more about that.

It’s just doing the math: what is the net revenue of a book? Take out the retailer fee, the distribution fee, the potential returns, the cost of goods—all of those expenses put together, you end up with a ratio that is somewhere between one employee per $500,000 and $1 million per employee. That is the optimal operating ratio. There are these incredibly wonderful publishing companies that have carved out niches that they are the dominant publisher and they have great revenue, but they’re operating at a loss, just because it’s so labor intensive. And they’re operating at a loss with employees who are making pretty awful incomes, you know? Especially living in New York as an editor making $60,000. That’s really brutal, and I would like to figure out a way to begin to escape that system.

The ratio definitely applies more the larger you get. So if you’re a $150 million company, having around 150 employees. Part of the economics of it is the larger you get, supposedly the more efficient you get, and you can actually push that ratio up to one employee per $1.5 or $2 million.

Back to Stable: I want to make sure I understand this, and readers understand this. If there’s a publisher under Stable, does Stable own that publisher outright?

So we have several joint ventures. With Michael Jacobs [of Galpón Press], we are essentially serving as his distributor and his pool of resources that he can use if he wants to. But the primary goal of Stable is to fully acquire publishers, whether that is cash upfront or an equity swap, so that publisher now owns a piece of a much larger organization, which is what we’re doing with Ulysses and Velo.

I was wondering if you’re going into debt to make the acquisitions.

No.

So what pays the bills for your businesses?

Backlist pays the bills. At Ulysses, 70 percent of our annual revenue every single year is from backlist titles, and that guides our entire acquisitions model. We are not looking for books that are going to have that classic hockey stick curve—one week they sell 20,000 copies and 10 weeks later, they’re selling 10 copies. We are always looking for those books that really hit an audience’s need, whether it is for information or entertainment, and that book is going to have legs for a very long time. To me, that’s the only way you can possibly set up a publishing company to grow organically, … by focusing all of your efforts on what is going to work for the long term, not for just this year.

The other thing about Ulysses, which very much guides my approach to publishing, is treat publishing as though it’s an investment portfolio, which in a lot of ways it is. The best publishers in the country are going to bat .300—now I’m really mixing my metaphors—but three of their books are going to succeed, and then seven of them are going to fail to earn out or be successful.

What I want to do as a publishing company is put ourselves in the best possible position to get that batting average a little bit higher than most of the other companies. And part of it is keeping a very diversified portfolio. So how do you position your backlist so you can take advantage of swings in the market, swings in different categories, and the somewhat predictable cycles to a lot of these categories, where, you know, eventually the master cleanse is going to come back, or the Mediterranean diet will be rediscovered for the 50th time, and there’ll be a new readership looking for those books.

Ulysses doesn’t have what I would consider to be a traditional submissions process. You use data to identify what books you want to do, and then you find the expert, or the author, to write it. Is this model threatened by AI?

I think that the model is pretty much over, actually, at least in the way that we approached it, and companies like Callisto approached it. I don’t think that it is necessarily AI that has done it, but AI is definitely a pain point, especially when you see copycat books that are being published before your own book that you’ve worked so hard producing.

AI also has the potential to solve some of those problems, in improving search and moving search beyond natural language processing and keyword-driven algorithms—to really understand the value that has been put into a book, and the response that it’s gotten from critics, reviewers, and customers. I think that is going to start to change things toward really valuing books that have had a lot of work put into them.

But we’re not there yet. We’re kind of at this muddy crossroads where AI is a huge threat to everybody trying to sell things on the internet. The reason that I feel like data-driven publishing is changing now is that, in the 2010s when we were really crushing it, you could just come up with an idea, identify an audience, and make a lot of money publishing a book specifically for that audience. But for me, it was always an arbitrage moment that was not going to last forever because everybody is looking at the same data. Not everybody can be playing Moneyball before it just stops working.

So we have definitely changed our approach significantly over the last six years to account for this thought that everybody is seeing the same data, everybody’s going to act on it. Then how do you produce books that have some sort of moat around them? That can be anything from having a top-notch author and expert in the area who can really drive sales and protect the value of that book, or a license project where you are basically using money to protect that book, or a very unique package and approach to a topic that is going to be hard to replicate. But it’s very cutthroat now with data-driven publishing.

It’s been a very subtle transformation in our acquisitions meetings, where it goes from “Prove to us that this audience exists, that we know how to reach that audience, that there is an expert that can write what needs to be written for this audience,” to “Tell us how we’re going to protect this book when it comes to market.”

OK, let’s move onto everyone’s favorite topic, AI. We touched on AI agents a little bit when you discussed the vision for Stable. When I heard you speak at the NYU Publishing Institute in January, I was struck by something you said—that you think publishing companies will eventually be run by AI agents handling tasks. The reception in the room after you said that was not, I’d say, enthusiastic.

No.

I imagine you get a lot of pushback on that.

Yes. So I think this is definitely already happening within certain subsets of publishing. I don’t think there is a single established house that has been in operation for ten years that is moving in that direction entirely, but you are going to see new publishers crop up that take that approach, in the same way that you have publishers who are saying, you know, I don’t care about the rest of the market, I’m just going to put this book on KDP and I’m going to make a lot of money because it is as close to printing money as you can get.

I don’t think that AI agents are going to be dominating publishing in the next ten years. I do think there will be publishers who figure it out and can operate a full publishing company [that way]. There will be a lot of advantages to using AI, which is very cheap now and is getting cheaper, to augment workforces. The thing that scares me the most is if, you remember that Publishers Weekly chart about the number of people working in publishing? I think that trend line continues to go down.

My hope is that by augmenting employees with AI that we can see wages start to go up, but it’s going to be harder and harder to get into the industry. As a CEO who has fiduciary responsibilities to our shareholders and my family, there is a part of me that is eager to take advantage of technology that can really revolutionize our productivity with a small team. But the publisher and the human in me is really sad about AI. I find it kind of heartbreaking, and it has taken a lot of the joy out of some of the tasks that I derived a lot of joy out of.

And I find it really sad that we’re at a point where you are never going to really know whether an author wrote the material they’re submitting or if they had AI help them. In the next couple of years, we are probably going to give up the title of the most intelligent species on the planet to something that we created. So I am not a techno-optimist. I am really sad that it is here, but I know that we have to figure out how to use it because we’re a business that competes with other businesses, and we’re an industry that competes with other entertainment industries and everybody is fighting to figure this out.

Last two questions. In another interview you said, “If you’re licking your chops at the growth of audiobooks, [I] would suggest treading carefully.” Can you explain why?

This moment in audiobooks reminds me a lot of the beginning of self-publishing and the beginning of podcasting. My worry is that the barriers to entry for audiobooks are already collapsing because of synthetic voice. What you’re going to see is the same price pressures that have hit every other format come into play in audio, and it’s going to be a massive proliferation of audiobooks, many of them not well-produced, many of them using synthetic voices. And you’re going to see it play out very similar to what we saw with self-publishing and podcasting, where it looked like an endless gold rush for a minute, and then got very, very saturated. At the end of that process, there are few winners that remain, and a lot of people have put a lot of work into products that never sold anything. So that is my concern with audio.

I am a huge fan of audiobooks. Especially for the generations that are coming up now, that may not want to read as much, just being able to get that content to them in a way that they’re already enjoying entertainment is really powerful and wonderful for the industry. But I think that we are going to see those same price pressures come into play.

The last question has to do with Amazon and AI. Authors and publishers alike have always been anxious about what Amazon might do next to affect discoverability. How do you think Amazon’s use of AI might affect authors and publishers? What might we need to prepare for?

Oh, love this question. So the biggest thing I was always scared of when AI came out, and I had my first “Oh, shit” moments with it, was the search experience. We have been in a world for 25 years where search is driven by [specific] factors and every company invested in SEO. That era is over. The thing that I would push every publisher and author on Amazon is—this is going to change very significantly over the next couple of years. Amazon is currently still using their A9 search algorithm, which is NLP [natural language processing] and keyword based. And all of the metadata work that every publisher has spent the last 10 years doing has been geared towards that search algorithm. And the A10 algorithm that is coming online now [see Rufus] is going to completely change the way that books are found. For a publisher like Ulysses, we have to now go back through 1,500 different titles and optimize for a new AI-driven world that we don’t yet understand and is constantly changing.

What I see happening—if you use Rufus, you know. You ask, “I’m looking for a great military history book about the Civil War.” It doesn’t give you an endless scroll of results and sponsored products upfront and peppered throughout. It gives you three results, and those three results are based on how well does Amazon move this product? So what is the turnover like for the inventory? What are reviewers saying about it? What is the content on the page? What does the cover look like?

All of these things can now be taken in through AI and weighted, and it gives you three results. I don’t know how you foster awareness and discoverability on Amazon, which has dominated 50 percent of the book market, in a world that no longer relies on the type of search that we’re used to. That, to me, is probably the single thing about AI that keeps me up at night the most.

Can you use AI to revise your book descriptions to meet this challenge?

That is part of what we want the AI agents to do with Perfect Bound, to go through your metadata and make adjustments based on keywords that are trending, because that will continue to be a part of it. But AI is going to know if AI wrote this passage about the book. That’s the whole thing about the future of SEO: ironically, it’s looking for human-made content to verify that this is a good product. So you have to tread very carefully with AI-generated content on your book descriptions and even with the materials. To me, it just looks like a lot of work is coming our way, and not a lot of time to make it happen.

Glad I could end the interview on that wonderfully optimistic note!

(laughs) No, I do want to end by saying I am more optimistic about publishing now than I have ever been, because we are a product that you do not have to read on your screen. And there is a phenomenal amount of hunger for well-researched, human-derived content and material, and that is not going to go away ever. There is so much value in this industry. We are now at a point where, instead of regretting [my] English literature degree right out of school, I am so proud and happy that that is the route that I took. It puts all of us in this industry in a really strong place going forward. We have fantastic critical thinking skills and communication skills.