How to Turn Your Expertise Into a Book

You have explained the same idea to clients for years: on a whiteboard, on a discovery call, at a conference after two glasses of wine. People nod, take notes, and ask whether you have written it down anywhere. You haven’t, or not properly.

When you try, the result is thinner than the thing in your head. The method comes out as a list of steps anyone could have written. The parts clients pay for, the moment you tell someone not to do the obvious thing, never reach the page.

That gap is the real problem. It is not a shortage of knowledge or words. Expertise is stored in a form you cannot easily read back to yourself, and the way you extract it decides what kind of book you end up with.

What expertise looks like once it is on the page

Readers do not buy “expertise”. They buy four things that, together, feel like it. Each is extracted differently, and most first drafts are missing at least one.

Method. The repeatable sequence: what you do first, second and third, and the tools at each stage. Most experts can already write this part, and competitors can most easily copy it.

Judgement. The decisions inside the method: when you skip a step, when you reverse the order, the signs that a client is not ready. Judgement is the layer most expert books lack, and the one a reader cannot get from a free course or a chatbot. Method without judgement reads like a programme handout.

Stories. Cases that show the method and the judgement working, or failing, on real people. Readers recognise themselves in stories, and most confidentiality problems live there.

Evidence. What makes a careful reader believe you: client results you have permission to describe, figures with a source, the reasoning behind a rule. AI tools are most likely to invent this layer when it is missing.

Self-test: take one chapter idea and write a line under each heading. Wherever you write “I’d have to think about that”, you have found the work.

Why writing down what you know leaves most of it out

In a Journal of Surgical Research study, Richard Clark and colleagues at the University of Southern California asked expert trauma surgeons to describe a procedure they perform routinely. Those who described it unaided “omitted nearly 70% of necessary decision steps”, while a structured interview with a single surgeon produced a much more complete account (Clark et al., 2012).

A follow-up in Academic Medicine filmed three experts teaching another procedure. They left out an average of 73% of decision steps. With structured prompts, the share of steps they described rose from 44% to 66% (Sullivan et al., 2014). The authors put it down to “the automated nature of knowledge”.

These are small surgical studies, not studies of coaches, but we think the pattern transfers. Miro Kazakoff, who teaches managerial communication at MIT Sloan, puts it plainly: “When we see a pattern or recognize something or know something, we forget what it was like before we knew that thing” (MIT Sloan).

Steps go missing too, but decisions go missing most, and decisions are your judgement. Whatever route you choose, something has to ask you the questions you would not ask yourself.

Four ways to get expertise out of your head

Every expert book comes out of one of these routes, or a mix. Each suits a different kind of thinker and fails in a different way.

Write it yourself

You sit down with a chapter plan and draft. This suits people who already write regularly (a weekly newsletter, reports clients actually read) and think best on the page. It gives the most controlled prose and the clearest argument.

It is also the slowest route, and the most exposed to the omission problem: alone, you explain the steps you are conscious of. We would not start here if you are a practising coach with a full diary, unless writing is already how you think.

Talk it out

You dictate, or transcribe talks you have already given. Speech is fast. In a Stanford-led lab study, speech input on a phone was 2.93 times faster than typing for short English messages, 153 against 52 words per minute (Ruan et al.). That is short messages on a touchscreen, not chapters, but the direction is clear.

The catch is that a transcript is not prose. Talk repeats itself, skips steps the audience could see on a slide, and leans on tone. Our guide to turning talks, webinars and sessions into a book covers the repair work, and the main site’s guide to turning podcasts and coaching calls into a book covers transcription.

Be interviewed

Someone else asks the questions: a ghostwriter, a book coach, a colleague, or software built to interview you. This is the route the surgical research favours, because a good interviewer keeps asking “what would make you do it differently?” until the judgement comes out.

It suits busy experts who know more than they can easily say. Quality depends entirely on the questions; a friend reading a generic list gets generic answers. Do the groundwork in our guide to extracting your knowledge before writing a book first, so interview time goes on decisions rather than biography.

Assemble it from what you have already made

You build the book from material that exists: workshop decks, proposals, newsletters, programme outlines. Most experienced coaches have far more of this than they think, and it is usually their best-tested thinking.

The risk is a stitched-together archive with no argument connecting the parts. Documents are strong on method and weak on judgement, because slides record what you teach, not why you deviate from it. The audit method is in writing a book with AI from your documents.

Choosing a route

Route	Suits you if	Strongest on	Typical failure
Write it	You already write weekly and think on the page	Argument, control of voice	Slow; judgement left implicit
Talk it	You speak fluently and have recorded talks	Stories, energy, speed	Transcript mistaken for a chapter
Be interviewed	You are busy and know more than you can easily say	Judgement, missing steps	Only as good as the questions
Assemble from documents	You have years of decks, posts and proposals	Method, tested frameworks	An archive with no argument

Our default: documents first, interview second, writing last

For most coaches and consultants with ten or more years of client work, we would combine routes in that order. Start with what you have already made: it costs little to extract and shows what is covered. Use an interview for the gaps, especially judgement and stories. Write last, when you are shaping an argument from material that exists.

Some authors disagree, fairly: writing is how they discover what they think. If that is you, write first and keep documents as a reference shelf.

If you want the documents-plus-interview sequence in one place, guided book software is one option. WritebookAI analyses PDF, Word and text files and pasted notes, shows what they cover and what is missing, then holds a verification interview about conflicts and gaps before producing a book strategy and chapter plan for you to approve. When a chapter lacks something, such as a client result or a statistic without a source, it leaves it out and asks you rather than inventing it. It is guided software, not a ghostwriting service; you verify every claim. If you want someone else to carry the whole process, a human ghostwriter is the better fit; we compare the options in writing a business book without a ghostwriter.

A worked example: one chapter, four layers

Imagine Priya, a hypothetical sales consultant who has spent fourteen years helping founder-led B2B firms hire their first salespeople. Her planned chapter is “Do not hire a salesperson until you can describe how you close”.

Layer	What she has	Where it lives	What is missing
Method	A four-stage readiness check	A workshop deck she has run twelve times	Nothing; it is well tested
Judgement	When she tells a founder to wait, and when to hire anyway	Only in her head	Almost all of it
Stories	A founder who hired a sales leader too early	Session notes and a podcast interview	Written consent, or a heavy disguise
Evidence	Her view on how long new hires take to become productive	A figure she quotes from memory	A source, or her own anonymised client data

The table makes the route obvious. The deck supplies the method. The judgement needs an interview, and the most useful question is something like “tell me about the last three times you told a founder not to hire yet, and the one time you were wrong”. The story needs a permission decision before it goes near a draft. The figure gets a source or comes out.

Priya’s instinct was to start by writing. She would probably have produced the four stages neatly and left out the part founders pay her for.

Five decisions to make before chapter one

First, a readiness check.

If two or more boxes are empty, the honest advice is to wait: run more client work, publish shorter pieces, and come back when the method has settled. If you pass, make these decisions in writing.

1. The reader. One person at one moment: the newly funded founder, the first-time manager, the CFO before an audit. Not “leaders”.

2. The promise. What is true for the reader after the book that was not before. Josh Bernoff, who co-ran a study of business-book returns, names unclear goals, no clear audience, no differentiation and no marketing as the reasons business books fail, and offers a test: can you truthfully finish the sentence “This is the first book that…”? (Bernoff). If not, keep narrowing.

3. The book’s job in your business. Leads, speaking invitations, better sales conversations, a door into a course. A book meant to start sales conversations needs a clear next step; one meant for the stage needs a quotable idea.

4. Scope, including what you give away. Our view: give the method away in full. Readers who can do it alone were never going to hire you, and those who try usually discover the judgement is the hard part. You hold back the doing, not the knowing. The main site’s guide to turning expertise into a book framework covers naming and structuring a framework, and our article on turning a framework into a book covers the chapters.

5. Permissions and who holds the pen. List every client story and decide how each will be cleared. The ICF Code of Ethics (effective 1 April 2025) defines confidentiality as protecting information obtained “in or around the coaching relationship”, lists the client’s written consent among the exceptions, and in standard 2.5 extends coaches’ obligations to the software and AI they use. Former employees should check what their contract says about proprietary material. If AI will draft text, Amazon’s KDP content guidelines, as of November 2026, require you to disclose AI-generated text, even if heavily edited, but not AI-assisted work where you created the text yourself.

Mistakes that keep expertise off the page

• Starting with the outline. An outline from memory reflects what you talk about most. Sort first; see organising years of experience into book chapters.
• Writing the method and calling it done. Chapters of steps with no decisions lack judgement.
• Asking AI to fill the gaps. A missing figure is information; a fluent paragraph hiding it is a liability. See stopping AI inventing facts in your book.
• Interviewing yourself with soft questions. “What do you believe about leadership?” gets a speech. “What did you tell the last client who wanted to skip stage two?” gets judgement.

Start with one client problem, not the whole book

Pick the client problem you solve most often. Record yourself explaining it for twenty minutes, as you would to a new client, then sort every sentence of the transcript into method, judgement, story or evidence. Expect the judgement column to be thin and the evidence column to hold figures you cannot yet source.

That tells you which route you need. If the method is all there, your documents will carry much of the book. If the judgement is missing, plan for an interview. Then make the five decisions in writing, and only after that open a chapter plan. More guides on each stage are collected under turning expertise into books.