The Edelman-LinkedIn B2B Thought Leadership Impact Report found that fewer than half of B2B decision-makers describe the thought leadership they consume as "good," and only 15% describe it as "very good." Most expert content does not clear the bar. Most expert books do not either. The problem is not that experts lack things to say. It is that most ideas that work in a 20-slide deck have never been tested across 70,000 words of continuous argument.
That is the test a book runs. Not the marketing test. The structural one. The real answer to why every expert should write a book lies in what that test reveals about the thinking, not in what the book does for the author's profile.
Marion's post on Tuesday made the inside-out version of this argument. She wrote that the first draft is the conversation in which the book tells you what it is about. That is true, and I have seen it happen. My argument here is one structural step higher. Before you find out what your book is about, you have to decide whether running the test is worth it at all. For most experts, it is the first time that a test will ever be run on their thinking.
The Real Question for Consultants and Thought Leaders
The cliché answer is "yes — your book is your best business card." That framing is exhausted, and it is also slightly dishonest. A mediocre book is not a credential. It is a liability. In 2023, Amazon introduced AI-disclosure rules on KDP in response to a flood of generated titles; the Authors Guild welcomed the move as a first step and called for consumer-facing labels — and the AI book glut has continued. Readers now meet generic expert books with suspicion by default. The bar quietly shifted. It moved from "I published a book" to "I published a book that proves I actually thought something through."
So the question is not "will a book attract clients." The honest question is: will the act of finishing a book make me a sharper version of the expert I already claim to be?
That is a different question. It does not depend on the book selling. It depends on whether the writing forces something.
What 70,000 Words Actually Do to Your Thinking
In conversation, an expert's idea has a lot of helpers. Tone of voice. The other person's nodding. The fact that you can change the subject when a thread starts to wobble. None of those are available on page 142.
A book does not let you change the subject. Every chapter has to remain consistent with every previous chapter and survive every chapter that comes after. The structure is load-bearing. If chapter three is a stress test of one of your claims, chapter eleven cannot quietly contradict it without the manuscript starting to creak.
The educational psychologists Carl Bereiter and Marlene Scardamalia drew the line between two kinds of writing decades ago — knowledge-telling and knowledge-transforming, in The Psychology of Written Composition (Lawrence Erlbaum Associates, 1987). One is writing down what you already think. The other is writing that changes what you think while you do it. Most experts have only ever done the first. They are fluent at it. They mistake fluency for clarity. The 70,000-word format, in my experience, is the first place that mistake gets reliably exposed — fluent local prose does not survive a global structure that is genuinely demanding.
Marion and I co-wrote The Intentional Workplace: Walk the HYBRID Way. We had spent months debating how to name and characterise the post-COVID processes reshaping organisations around the world. We started with an acronym we wanted to leave behind as a mnemonic. The writing forced us to drop that and rebuild from data — each of us took three to four companies, read every story they had published about how they switched to hybrid work, and validated the claims against what those companies said had actually made them successful. We came back together, cross-checked the patterns, and refined the methodology and the narrative based on the language other practitioners were already using. The acronym we landed on was HYBRID. The same letters we started with. They survived only because the data happened to fit. That is the test the writing forced — and it is not a test any of our pre-writing conversations could have run.
That is the part most experts underestimate. They think they are writing the methodology they already have. The writing rebuilds the methodology from the inside.
The Counterargument: "Just Write a Shorter Piece"
The honest objection is not "use a ghostwriter." Most experts are not weighing a book against €6,000–€40,000 in nonfiction ghostwriting fees — and most do not even know that is the range, with bestseller-credentialed ghostwriters charging far more. The real objection is the cheaper one. I could just write a long article. A LinkedIn series. A white paper. A keynote. Why a book?
Each of those formats is good for what it is. None runs the same test.
A 1,500-word article asks you to land one sharp idea. You can do that without ever pressure-testing whether it sits inside a coherent system of ideas. A LinkedIn series rewards punch over consistency — readers meet each post fresh and forget last week's. A white paper has a fixed audience and a known outcome; it does not have to surprise anyone, including its author. A keynote can be brilliant and still be 40 minutes of the same idea repeated with better lighting.
A book is the only format where you have to hold an argument together across roughly 70,000 words without losing yourself, your reader, or the thread. That length is not arbitrary. It is the point at which the dependency graph between your own claims becomes too long to keep entirely in your head. You start needing the page to remember for you. And once the page is remembering for you, you can finally read your own thinking from the outside and notice where it does not hold.
Shorter formats let you avoid that. The book does not.
Why the Bar Shifted in 2025–2026
Two things happened at once.
The first is that AI-generated text became fluent enough at expert tone that platforms had to start filtering it. The second is that readers — especially the B2B decision-makers of the Edelman-LinkedIn report surveys each year — began reacting against generic expert content faster than they used to. The same survey that found fewer than half of decision-makers rate the thought leadership they consume as good also found that readers are increasingly able to spot recycled, shapeless arguments. They have had a lot of practice.
The signal a credible book sends now is no longer "I published one." A pile of AI-generated books was published this year alone. The signal is what a generated summary cannot fake: a coherent, specific, lived argument carried from premise to conclusion, with the author visibly being the one who held it together.
That is a higher bar. It is also a more honest one.
What This Means Practically
Not every expert should write a book. Not every idea is ready. The discipline of finding out which ideas are ready — and which experts are willing to do the rebuilding the format will demand — is itself part of the value.
Most of the consultants I work with through my-book.ai discover something in chapter four that means chapters one through three need restructuring. That is not a setback. That is the test working — the same load-bearing-system test I described last week in the codebase framing. One of our current client engagements is The PEARL Method: Leading Productivity that Lasts by Martin Kearns, due out shortly. It is an example of the kind of book the format rewards: a methodology its author has practised for years, finally being made to hold itself together as a system rather than a deck.
A book is a system. It has components, dependencies, and failure modes. Treat it like one and 70,000 words becomes manageable. Treat it like a long article and the structure will quietly sag two-thirds of the way through.
The consultants who come out the other end of the process are sharper practitioners. Not because the book attracted clients — though it often does — but because finishing it forced them to find out what they actually believed. Most of them did not know, with structural precision, before they started.
I have never published a solo book. All four of mine are co-authored — Agile Transition: What You Need to Know Before Starting (2012), The Hitchhiker's Guide to Agile Coaching (2017), ORGANIC agility Foundations: Leadership and Organization (2020) with Dave Snowden and Geoff Watts, and The Intentional Workplace: Walk the HYBRID Way (2025) with Marion. That is not a footnote. The format is hard enough that being held accountable by another author, or by a writing process that does not let you cut corners, is what makes the test honest. The point is not the solo heroism of finishing alone. The point is being held to the structure long enough for the thinking to become load-bearing.
If This Sounds Useful
If you have a methodology you have been practising for years and you suspect — but have not confirmed — that it would hold across the length of a book, that suspicion is the right place to start. The work is in finding out. The book is the evidence that the finding-out happened.
If that kind of collaboration sounds useful, I'd like to hear about your book.
Andrea Tomasini is the founder of my-book.ai and co-founder of agile42. He has co-authored four books across software engineering and organisational design, including The Intentional Workplace: Walk the HYBRID Way and ORGANIC agility Foundations: Leadership and Organization. He spent two years engineering the writing system that powers my-book.ai after watching the first book through it surface every consistency problem long-form writing produces and large language models cannot solve alone.