Some reflections on writing and publishing my book 'AI-Powered UX research' and on non-fiction tech publishing

I am writing this on a plane to the Outliers conference, where I am giving a talk about my book, signing copies, and taking questions. Somewhere after takeoff I started asking whether the book has been a success, and I do not know. AI-Powered UX Research came out on July 15, 2026. It sells between one and twenty-five copies a day, every day. That is not retirement money, but it is strangers paying for a thing I wrote on my own evenings, and I will take it.

What made me ask is that a couple of publishers have offered me an advance to write a second one. That got me thinking about the first time around, when I talked to a handful of publishers and said no to all of them. This is the post where I explain why, the way I would explain it if you were sitting next to me on this flight and asked. So it is going to be a bit of a rant. You were warned.

They wanted eighteen months for a book about tools that change every quarter

Every publisher I talked to quoted a year to eighteen months from a signed contract to a book in stores. That is for a book about AI tools.

Think about what eighteen months means for this subject. The tools I used while writing got replaced twice before I finished the manuscript, and that was during the writing, not after. Add another year and a half and you are paying $22 content that is wildly out of date.

I understand why they are slow. Publishers plan in seasons, the sales reps sell each season to bookstores months ahead, and the printing has to be scheduled. It is a whole machine, and the machine runs at the speed of paper. That is fine if you wrote a book about leadership. Nobody's leadership book goes stale. Mine would have been a historical document by launch day.

They wanted the rights until 70 years after I die

This is the part where I lost my patience. The contracts asked for every format, every territory, every language, anything derived from the book (a course, a workbook, a talk based on it), and an option on my next book. An option means they get first look at whatever I write next, which is a bold thing to ask for from someone whose first book you have not sold yet.

Then I read the term. Most publishing contracts grant those rights for the full term of copyright, which in the US is the life of the author plus 70 years. Seventy years after I am dead. My grandchildren would be negotiating with these people.

I want you to sit with that for a second, because I did. What do you mean, forever? I wrote the thing on evenings and weekends, on my own time, and the offer is that a company that has never sold a copy of anything I wrote owns it into the next century, in exchange for a printing schedule and a line in a catalog.

There is supposed to be an escape hatch. If a book stops selling, the out-of-print clause lets you ask for the rights back. Except that in 2007 Simon & Schuster rewrote its contract so a book does not count as out of print as long as it is available in any edition, including electronic. With print-on-demand every book is available forever, so nothing ever goes out of print and nothing ever comes back. The Authors Guild called it a rights grab at the time. That was 2007. Draw your own conclusions about what the contracts look like now.

And for all of that, they pay you a single-digit percentage of the cover price on a paperback. On a $22 book that is less than a subway ride.

The publishers have no idea what sells

My favorite part of this whole story happened before I ever talked to a publisher, in a courtroom in 2022. The US government sued to stop Penguin Random House from buying Simon & Schuster, and the trial put the CEOs under oath for three weeks.

Markus Dohle, who ran Penguin Random House at the time, explained the company's name to the court like this: "Everything is random in publishing. Success is random. Bestsellers are random." He said that under oath, and he ran the biggest publisher in the world. Testimony from the same company put the share of its books that make a profit at 35 percent.

Jonathan Karp, the CEO of Simon & Schuster, testified that publishers do not guarantee a marketing budget, and compared promising an author success to taking credit for the weather. He also called a $100,000 advance fairly small, which will be news to nearly every author alive.

So put it together. The pitch is that they know what sells and they can make it sell. Their own CEOs, under oath, said they cannot tell what sells and will not promise to try. That is the deal on the table, and that is what they want the rights for your life plus 70 years in exchange for.

They wanted my audience and would not promise to spend a dollar on marketing

Every nonfiction proposal has a section called platform. Platform means how many people already listen to you: subscribers, followers, people who show up to your talks, with numbers attached. If you do not have one, the proposal usually does not get read.

So they want you to arrive with the readers. Then, per their own testimony, they will not commit to a marketing budget, and in practice most of the promotion lands on you anyway. You bring the audience, you do the marketing, and they own the book. I kept turning that over trying to find the part where the publisher does something, and the part I found was the calendar.

My book sells about 95 percent in the US, almost all of it on Amazon. It is also on Barnes & Noble, Books-A-Million, and Walmart online, which I set up myself. Nobody has bought it by finding it on a shelf. The people buying it, the one to twenty-five a day, come from this blog and from LinkedIn, and they were already here before I talked to anyone.

I am not saying a publisher gives you nothing. They give you editing, design, bookstores, libraries, foreign rights, sometimes an advance, and a logo on the spine. For some books that is the whole point. For a practitioner book that people read on their phone, it was a list of things I did not need, priced at everything I had.

Doing it myself cost $6,500, and every paperback pays me $9.20 instead of $1.65

The bill was simple. Design was $3,000 for the cover and the interior. Copyediting was $3,000. Another $500 or so went to design-adjacent things, the audiobook cover, the LinkedIn banner, the images for the Amazon product page. Six and a half thousand dollars all in, not counting ads and not counting the audiobook, which went through ACX and is its own story.

The part nobody puts in the LinkedIn post is what a copy is worth. The paperback is $22. Amazon's Kindle Direct Publishing pays 60 percent of the list price minus the cost of printing the copy, and Amazon's own example prices a 250-page black-and-white paperback at $4.00 to print. So $13.20 minus $4.00, which is $9.20 a copy, paid to me.

A standard publishing contract pays 6 to 7.5 percent of the cover price on a paperback. Take the top of that range and a $22 book pays the author $1.65 a copy. That is the subway ride from earlier.

So my $6,500 comes back after 707 copies. Under a contract the same 707 copies would have paid me $1,167, and every copy after that pays $1.65 instead of $9.20. At one to twenty-five copies a day you can do the rest of that math yourself, and I have.

Ebooks are worse. Amazon pays 70 percent on an ebook priced between $2.99 and $9.99. A publisher pays 25 percent of whatever is left after the store takes its cut, which means the store gets paid, then the publisher, then you, in that order.

And at the end of it, the book is mine. Every one of those one to twenty-five copies a day pays me, not a company that would have shipped it eighteen months late. If it stops selling I can drop the price, rewrite a chapter, give it away, or pull it, and I do not have to ask anyone. Nobody was offering an advance on the first book that changed any of that. If someone had, this would be a shorter and more grateful post.

The book was the smaller job

Writing the book was the part I knew how to do. Everything after the last chapter was a second job I had not applied for: publisher, art director, ad buyer, distributor, and the person who explains to an Amazon form why the ISBN is fine.

The publishers' pitch is that they take that second job off your hands. Their own testimony says they cannot pick winners and will not promise to market the book. The proposal template makes the audience your job. The contract keeps the rights until 70 years after you die. I do not have a polite word for that pitch.

The people are fine, and I mean that. The business is a machine built for bookstores in another century. The parts of it that protected the publisher survived, and the parts that helped the book got handed to the author, with the invoice.

The one thing I would pay for again, and did, is the editor. A copyeditor who pushes back on every third sentence is worth the $3,000, and it is the only line in the budget I never questioned.

If you are writing nonfiction and someone offers you a deal, I am not telling you to walk. If you need airport bookstores, a publisher can get you there and you cannot. If you have no audience, their catalog is the only distribution you will get, thin as it is. If your book will still be true in five years, the eighteen months costs you nothing.

But read the term, read the out-of-print definition, read the option clause, and then ask them what they will spend on marketing, in dollars, in writing. Then decide. The first time, it was not worth it to me.

Which brings me back to the offers. A couple of publishers want to pay me an advance for the second book. I am flattered, and I am also reading the same contract with a check attached to the front of it. The calendar has not changed. The term has not changed. The audience is still the one I brought. I have not decided, and the fact that I have not decided tells you the money is doing some work on me. Ask me at the conference.

If you want more of this, subscribe at thevoiceofuser.com. If you want the book, it is on Amazon: AI-Powered UX Research: How to Run Research at the Speed Your Team Actually Needs.