What Does a Startup Coach Actually Do?

Ask ten founders what a startup coach does and you will get ten answers: some say strategy, some say therapy, some say expensive accountability. All three are partly right, which is exactly why the category is easy to buy badly.

What happens in an actual session

A coaching session is mostly questions. You arrive with a situation — the co-founder conversation you keep postponing, the pricing decision you have re-made four times, the growth that stalled — and the coach spends most of the hour making your own reasoning visible to you.

• They separate the problem from the story. "Sales are slow" usually turns out to be "I have not asked anyone to buy in three weeks."
• They surface the assumption you did not know you had. Most stuck decisions are stuck on one unexamined belief.
• They make you commit to something specific. Not "improve the pitch" but "run it past three people before Friday."
• They follow up. The follow-up is half the value; without it, coaching is a nice conversation.

Coach, mentor, consultant, therapist

• Coach — asks. Improves your decision-making. Usually paid, usually structured, usually time-bound.
• Mentor — tells. Shares what worked for them. Often free, often domain-specific, sometimes wrong for your context.
• Consultant — does. Delivers an answer or an artefact. Right when you have a defined gap, wrong when the gap is you.
• Therapist — heals. If the block is anxiety, burnout or grief, coaching is the wrong tool and a good coach will say so.

What a startup coach cannot do

They cannot make the decision, cannot know your market better than the customers do, and cannot substitute for evidence. If you have not spoken to ten customers yet, hiring a coach buys you a well-facilitated conversation about a hypothesis nobody has tested. Go get the data first.

What it costs, roughly

Rates swing wildly by market and seniority — group programmes and peer circles sit at the accessible end, one-to-one work with an experienced operator at the top. Before you commit, ask: how many founders at my stage have you worked with, what does a session look like, and what happens if it is not working after three sessions. The third question separates professionals from personalities.

Do you need one yet?

Two honest tests. First: is your bottleneck information or behaviour? If it is information (do people want this, what should we charge), go and find out — that is validation, not coaching. If it is behaviour (I know what to do and I am not doing it), coaching helps a lot.

Second: have you had any real reps recently? Most of what a first coaching engagement uncovers is practice you never took — pitches never said out loud, sales conversations never had, difficult questions never faced. That part you can do now, for free.

Practice first, then coach

Founders who arrive at coaching with data get far more out of it. Run your pitch past an AI investor panel, take a real sales conversation in sales roleplay, or talk the whole thing through with an AI coach that pushes back from four different angles. Then, when you hire a human, you are paying for judgement instead of paying them to watch you discover the obvious. More on choosing one in our guide to founder coaching.