Is coaching right for me? Two ways to explore

You can try a coaching-style conversation before committing to training. Use it to notice what you did, how you felt and what you want to learn. One conversation cannot establish whether coaching is the right career for you.

There are two independent starting points: a fictional conversation where an AI plays the client, or a practice conversation with a willing adult who knows you're a beginner. Neither requires trying the other first.

• Explore with a fictional AI client if you want to try asking and listening without using anyone's real situation. Check the tool's charges, access conditions and data settings.
• Explore with a willing adult if you want to hear a real agenda and the person's account of the conversation. Disclose your lack of training and agree the limits before starting.

What you'd be trying

The National Careers Service says life coaching is not regulated in the UK. That describes the occupation's regulatory status. It does not establish competence, safety or legal clearance for particular work, including specialist, clinical or overseas work.

The Skills England coaching professional standard, updated 14 May 2026, describes listening and reflective questioning alongside contracting, boundaries, confidentiality, ethics, record keeping, referral and supervision.

In an exploratory conversation, notice whether you listen, ask rather than tell, and stay with someone else's thinking instead of jumping to your own answer. You may also encounter boundaries or the need to stop. A single conversation cannot establish competence in those responsibilities.

Try a fictional client played by AI

You take the coach's role. The AI supplies the fictional client's replies rather than coaching you. Choose an invented person with an everyday, low-stakes issue, and keep real names, employers and private details out of the conversation, including your own. Assume anything you type may be stored by the tool you're using.

A simulated client has no real stake in the outcome. You can rehearse discussing boundaries, but its agreement does not establish informed consent or show how you would handle a real safeguarding concern. Do not describe the roleplay as coaching delivered to a real client. For any credential, check the body's own requirements rather than treating simulation and client delivery as interchangeable.

An illustrative prompt to adapt

Model behaviour varies between tools and versions. Fill in the bracketed fields with an invented person; do not model them on a colleague, friend or client.

You are playing a fictional client in a coaching roleplay. I am the coach and I am a complete beginner.

Client: [first name], [age], [job or life situation].

The situation they bring: [one everyday, low-stakes issue, for example deciding whether to apply for an internal promotion].

How they behave: they are not sure what they want, they answer briefly at first, and they do not hand me a neat summary of their problem.

Rules for you:

• Stay in the client role. Do not coach me, evaluate me or offer tips while we are in the conversation.
• Keep your replies to two or three sentences.
• Nothing distressing, clinical or in crisis. If I take the conversation there, say the roleplay is stopping.

Begin by telling me, in your own words, what is on your mind.

Choose a short duration, such as ten or fifteen minutes, and stop at the time you chose. This is an exercise suggestion, not a required practice dose. Stop sooner if the conversation becomes distressing or moves outside the agreed topic; don't rely on the model to enforce the prompt's stop rule.

If the AI starts giving you feedback mid-conversation, ask it to continue as the client. Persistent role changes are a limit of the tool, not an assessment of you.

While you're in it, notice whether you start proposing solutions instead of asking about the fictional client's situation. Write down an example afterwards. When time is up, say the roleplay has finished. Reflect on your own before asking the AI for observations, so you have your own account to compare with its generated critique.

Treat that critique as prompts for further thought. AI praise does not establish whether you would be a safe or effective coach.

Try a conversation with a willing adult

Ask an adult who can comfortably decline and who isn't dependent on you or under your authority. Say plainly that you have no training and are exploring rather than offering a professional service. Agree a low-stakes topic, an end time, confidentiality limits, notes and recording choices, and either person's right to stop.

This brings a real agenda you didn't script, a reaction to notice in the moment and the person's account afterwards. Their feedback is limited to their experience; your relationship may affect what they say. They are not being asked to assess your competence.

Keep clinical, high-risk and specialist-advice topics outside the exercise. If someone is in distress, stop practising and help them seek appropriate support. In England, the NHS advice on suicidal thoughts points to 111 or 111 online for support. If someone has seriously harmed themselves or feels about to do so, call 999 or go to A&E. Elsewhere, use local crisis or emergency services.

Use the human practice guide for the invitation, full agreement and end-to-end exercise. You can choose this route without doing an AI roleplay.

What the research can tell you

The CARE study, published at CHI 2026, randomised 94 participants into text-based practice with an AI-simulated patient, with or without structured AI feedback, during a single 75-minute session. The authors describe them as novice counsellors recruited through Prolific. They were not a cohort of counselling students: reported backgrounds varied, only 10% selected counselling as a field, and 22.4% reported no formal education in the listed fields.

The study used automated scoring of counselling microskills. Within the practice-and-feedback group, reflections and questions improved significantly from before to after practice; empathy and suggestion changes were not statistically significant. Within the practice-only group, inappropriate suggestions decreased significantly, empathy scores worsened significantly, and reflections and questions showed no statistically significant improvement.

Those within-group findings do not establish that feedback outperformed practice alone on every measure. Comparing the groups' changes, the empathy difference was statistically significant; the differences for reflections, questions and suggestions were not.

A 2025 study of counsellor trainees analysed nine mock session transcripts and held a focus group with four of those participants. These students had already completed counselling-skills teaching. They described a low-pressure, repeatable opportunity to practise, with reservations about how agreeable the simulated client was. These were reported experiences, not measured coaching competence.

A reflection worksheet

Complete this soon after the conversation, while you can still remember specific moments. Copy the six blocks into a private note or document. Leave the other person's name and identifying details out, including when describing their topic.

This worksheet asks for examples rather than numerical ratings because it is not a validated assessment. An observation such as “I interrupted twice in the first five minutes, both times to suggest something” gives you a specific occurrence to revisit. There is no total, weighting or pass mark.

The conversation

• Date: [fill in]
• Who played the client: [fictional AI client / willing adult who knows I'm a beginner]
• Roughly how long: [fill in]
• Topic they brought, without identifying details: [fill in]
• Was I nervous beforehand? What about? [fill in]

What I did

• One question I asked that changed the direction of the conversation: [fill in]
• One question that only gathered information: [fill in]
• A moment I gave advice or steered towards my own answer; what was said and what I think prompted it: [fill in]
• The longest silence I let stand, and what I did with it: [fill in]
• Something they said that I did not follow up, and why I think I missed it: [fill in]

What I noticed in myself

• Where in the conversation did I feel most comfortable? [fill in]
• Where did I want it to stop? [fill in]
• Was I curious about this person's situation, or waiting for my turn? [fill in]
• Did I want to know how it turned out for them? [fill in]
• At the end, did I want to do it again? Straight away, next week, or not at all? [fill in]

What the other person said

Only complete this block if a human took part. Leave it blank for an AI client; keep any generated observations separately.

• What they said the conversation was like for them: [fill in]
• Anything they found unhelpful or uncomfortable: [fill in]
• What I am unsure I understood, without guessing what they thought: [fill in]

What I'd change

• One thing I'd do differently next time: [fill in]
• One thing I want to understand better before trying again: [fill in]

What happens next

• The next bounded thing I'll try, if any: [one conversation, one course taster, or a conversation with a working coach about their week]
• When: [fill in]
• What would make me decide this is not for me: [fill in]
• What would make me want to commit to training: [fill in]

Reading your notes back

A note about an interruption tells you what happened in that conversation; it does not establish a habit. A note that you kept thinking about the person's situation the next day gives you something to consider about your interest. Boredom may relate to the topic or your nerves rather than coaching itself.

An overall impression of how well it went can carry information, but cannot establish aptitude or competence on its own. The same limit applies if someone observed you: consider what they saw and the basis of their feedback rather than treating their reaction as a career verdict.

For deciding what to explore next, look at whether you wanted to repeat the exercise and what you wanted to understand better. This is a reflection suggestion, not a validated weighting of the answers. If you repeat it, write a new sheet before comparing; no particular number of sheets validates a conclusion.

You could try another fictional scenario, arrange a human conversation, seek observed feedback in a course taster, or ask your employer about learning to use coaching skills in your current role. You can also stop without buying a course.

For help choosing, read You enjoyed trying coaching. What now? The UK route map covers the wider training decisions.