Table of Contents
TL;DR: The worry is that a free tool made the paid human pointless. The math says the opposite. The 2024 Business Book ROI study found ghostwritten books returned a median of $92,500 and were four times more profitable than self-written ones, and the money came not from book sales but from the speaking, consulting, and credibility the book unlocked. AI touched none of that. It made clean prose free, which means the part you actually pay a human for, the part that earns the return, is now the only part that matters and the only part still worth paying for.
The worry, said plainly
ChatGPT is free. A ghostwriter costs thousands. Put those two facts next to each other and the conclusion looks obvious: why would anyone still pay for the human? It is a fair question, and the answer is not loyalty to tradition. The answer is in what you were paying for in the first place, which is a question almost nobody stops to ask before they conclude the human is now optional.
Before you make the cheap decision that costs you the book, walk through what the fee actually buys, what the numbers actually show, and what AI actually changed. The honest accounting comes out the opposite way from the obvious one.
What you were actually buying
You were never paying a ghostwriter to type. The typing was the cheap part even before AI existed. What the fee bought was judgment about what your book should be, the skill to pull the real story out of you, a voice held steady across three hundred pages, and the plain fact of the thing getting finished instead of dying in a drawer. None of that was ever about the words appearing on the screen.
The book that earns its keep is the book that says something specific, in a recognizable voice, that opens doors after publication. A book that does nothing reads competent and forgettable, which is precisely what happens when nobody is making the hard calls about what to include and what to cut. The fee paid for the calls, not the keystrokes. The calls were always the expensive part, and they still are.
What the numbers actually show
The 2024 Business Book ROI study surveyed 301 published authors and looked at where the money actually came from. Ghostwritten books returned a median of $92,500 and ran four times more profitable than books the authors wrote themselves. Here is the part that matters most: that money rarely came from selling copies. It came from what the book opened up. The median author pulled in $30,000 in new speaking work, $50,000 in consulting, $40,000 in workshops, and $64,000 in organizational sales after publication. Almost nine in ten said publishing was worth it.
My own clients track the same way. One Fortune 50 executive used his book to raise more than $30 million in venture funding and launch his own company. A Canadian entrepreneur turned his into a TEDx invitation. A LinkedIn branding book hit 15,000 copies in three days and reached number 43 across all of Kindle. Not one of those returns came from elegant sentences. They came from the doors a credible book opened, and a credible book required somebody making the hard editorial calls a free chatbot is structurally unable to make. The full list of these outcomes is on the case studies page if you want to see what the pattern looks like across different professions.
What AI changed, and what it left alone
AI made clean, grammatical prose free. That is real, and it genuinely ended a certain kind of writing work. The freelancer whose offer was “I will turn your notes into readable copy” is in serious trouble. The market has noticed, and rates for commodity prose have collapsed. If that was the part of ghostwriting you assumed you were paying for, the worry is warranted.
But look back at where the return actually lived. It lived in judgment, in story, in voice, in credibility, in the book getting done at a quality somebody would actually invite onto a stage. The machine reached none of those, and it cannot, because it is trained on the average of everything ever written and produces the average back to you. Average does not get you on a TEDx stage. It does not raise $30 million. What it gets you is a competent manuscript that does nothing. I made the longer version of this argument in the cornerstone piece on whether AI can write your book, and the short version is that AI automated the one part of the job that was never where the value was.
So, is it worth it?
More than before, not less. The cheap part of the work got automated, which leaves the valuable part standing alone and easier to see. You are no longer paying for words you could now get for nothing. You are paying for the exact thing that produced the $92,500 and the funding rounds and the TEDx stage, and that thing is more scarce now, not less, because the floor of competent prose dropped to zero and the gap between average and worth-publishing got wider in the same move.
If the book is worth doing, the case for doing it with a real human got stronger the day the tool went free. The version of the question that ends in “should I just use ChatGPT” stops being a smart cost-saving move and becomes a guarantee of producing something average, in a market where average is now so plentiful it is invisible. The version of the question that ends in “is the return there” is the one that actually matters, and for a book aimed at clients, audiences, or credibility, the return is there by a wide margin.
The middle option you may not have considered
There is a third path worth knowing about. If full ghostwriting feels like more than the project warrants but you also know a do-it-yourself AI draft will read flat, an AI-assisted book runs at roughly half the per-word cost. The machine carries the work that is not your voice, the research, the transcripts, the structure, the dull connective sections. A human stays on the voice, the stories, and the judgment, every time. That arrangement keeps the part that earns the return and trims the part that does not, and it is the answer for most authors whose worry is the price tag rather than the ambition.