How do I choose a CEO coach?

Choose on three things: whether they understand your situation without a tutorial, whether they have any stake in how you decide, and whether they are willing to be unpopular with you. Test all three in the first conversation. Then take a paid three-session trial before committing to anything longer.

Most founders choose badly because they run the process backwards, starting with websites instead of with their own problem.

Start with your problem, not with candidates

Before you speak to anyone, write down the sentence you have not said to anybody. The real one, not the tidied version. About your co-founder, the raise, or whether you want this in two years.

This does two jobs. It tells you whether you have a coaching problem at all, and it gives you the thing to test candidates against. If you cannot write it, you are not ready to hire, and the first three conversations will be vague.

Then sort what you wrote. If it is mostly "I do not know how this works," you want a mentor, cheaper and faster. If it is "I know what to do and I am not doing it," that is coaching territory.

Where to look

Other founders at your stage. The highest-yield source by a wide margin. Ask specifically what changed in how they operate, not whether they liked the person.

Your chair or a board member. Often better connected than they let on, and it removes any worry about how the board will read it.

Someone whose writing you already trust. You can hear how a person thinks before you pay them. Slow, and it rarely misleads you.

Directories and marketplaces will give you volume and no signal. They are a starting list, not a shortlist. There is more on sourcing in how do I find a startup coach.

The three questions

These separate the field faster than anything on a website.

"Describe a client you failed."
Anyone who has done this seriously has one. A coach who cannot name a failure is either new or managing you. Listen for whether they take responsibility or blame the client's commitment. The second answer is what they will say about you.

"What happens between sessions?"
If the answer is nothing, you are buying a fortnightly conversation. In my practice you write between sessions, and most of the change happens there rather than during the hour.

"Who is this not for?"
A real practice has turned people away. Anyone who says everyone benefits is selling. I do not work with idea-stage founders or people wanting general life advice, because the work assumes a company with real weight in it.

Watch for the stake

The structural advantage of a coach is that they have nothing to gain from your answer. That gets compromised in ways people do not flag.

Equity. A coach with shares now has a preference about whether you sell, when you raise, and how hard you push. Avoid.

Board-appointed and board-paid. They have an interest in you staying in the role. Fine for some work, poor for the question of whether you still want the job.

Company-funded generally. There is usually a quiet ceiling on what can be discussed when the payer has an interest in the outcome. If you can pay personally, the conversation has no ceiling.

Notice whether they disagree with you

The whole value is somebody who does not need you to like them.

If they spend the intro call agreeing, you have just watched the next six months. A good coach will push back on something before you have paid them anything, and will do it without being unpleasant.

This is also the moment to check comprehension. Describe something genuinely complicated from your last quarter. Someone with the right background asks their second question quickly and it is a good one. Someone without it needs ten minutes of setup.

Run a trial

Three sessions, paid, with a clean exit. A good coach offers this before you ask. Anyone making the exit awkward is showing you how the ending will go.

During the trial, look for four things.

You said something you have not said elsewhere. They asked a question that changed how you saw the situation. Something was agreed that you actually did. You left at least one session slightly annoyed.

Four out of four is a strong signal. Two is worth another month. Zero means stop, regardless of how much you liked them.

Settle the logistics before you sign

Cadence. Every two weeks is the working default. Monthly decays into a catch-up because too much happens in four weeks.

Length of engagement. Six to twelve months. Shorter gives you insight without change. Much longer with no clear reason and you may be buying company.

Crisis access. What happens when something breaks on a Tuesday. Some coaches offer it, some deliberately do not, and both are defensible. You want to know which.

Confidentiality, in writing. Particularly if a company is paying. Agree what, if anything, goes back to anyone.

How it ends. Notice period and what a clean exit looks like. Decide this while everyone is optimistic.

What to ignore

Certifications tell you someone completed a programme, which is why what actually makes a good CEO coach is a separate question. My view is that lived experience beats credentials here, and I will push back on anyone treating a certificate as the bar. Named frameworks are mostly repackaged versions of a handful of models. Client logos tell you a big name once bought a session. Price sorts poorly for quality in this market, in both directions.

Common questions

How many should I talk to?
Three is usually enough and gives you a comparison. More than five and you are avoiding the decision, which is worth noticing on its own.

Does my coach need to know my industry?
Rarely. Judgment, relationships and role questions are not sector-specific. Industry knowledge matters for advisory work, which is a different purchase.

What if my board wants to pick?
Reasonable of them to have a view and worth being clear about who pays and what is confidential. If you cannot get a private relationship out of it, consider paying yourself.

How do I end it if it stops working?
Directly, sooner than feels comfortable. How they take that conversation is the final piece of evidence about whether they were any good.

The short version

Write your real problem down first. Ask about a failure, ask what happens between sessions, and check whether they will disagree with you for free. Then trial it before committing.

More on the role in what is a founder coach, and on quality signals in what makes a good CEO coach. What I offer, and what it costs, is on the services page.