Coaching conversations

A coaching conversation helps a person think through their own goal and choose their own next step. You ask more than you tell. It differs from telling (you give the instruction), from mentoring (you share your own experience) and from advice in the form of a question ("have you thought about a course?").

Coaching fits when the goal is the person's and they know their work better than you do. When Olha says she wants to be ready to lead a team one day, the instinct is to hand her a list: run the Friday call, plan the autumn rota. But then it's your plan, and she'll follow it rather than own it.

The conversation has four stops: the goal, today's reality, the options, and one next step. The opening line should invite her to define the goal in her own terms: "What would 'ready to lead' look like for you by December?" A verdict ("you're not quite there yet") does the opposite: it puts her on the defensive before she has said what she wants.

The first two stops set up everything else.

The goal should be the person's own, concrete, dated and within their control. "Get better at leadership" is too vague to know when it's reached; "be promoted by October" depends on others; "run the Friday call" is the lead's idea. Olha's answer works: plan the September restart on her own by December (the rota, onboarding for anyone new, the hand-up rules), with the lead only reviewing.

Today's reality is about facts and what has been tried, not potential. Ask: "What have you already done on your own, and what got in the way?" Olha planned the hard-ticket rotation in February; what gets in the way is that when a newcomer is stuck, she takes the ticket because it's faster.

Avoid "why" questions here ("why do you do that?"): they sound like an interrogation and invite excuses. "What" and "how" questions keep the person describing, which is what this stop needs.

In the reality stop, you'll often spot a pattern before the person does. With Olha it's obvious to you: taking a stuck newcomer's ticket because it's faster is exactly your own January habit with the hardest tickets. The temptation is to say so.

Don't name it for her. Ask an open question about the consequence instead: "When you take the ticket, what does the newcomer learn?" Her answer ("That Olha fixes it. Hm.") is worth more than your explanation, because what people discover themselves they're far more likely to act on.

Three replies spoil this stop. Naming the pattern ("that's what I did in January, don't repeat it") turns coaching into advice. A "why" question about character ("why are you so impatient?") labels the person instead of looking at a behavior. And jumping to a rule ("nobody takes over a ticket") skips her options entirely and makes the solution yours.

Once the goal and the reality are clear, ask for options: "What could you try?" Then, whatever the answer, ask "What else?" at least twice. The first idea is usually the obvious one; the third is often the useful one, because the person has to step past their first instinct.

Olha's list grew that way: pair with the newcomer instead of taking over; write up the five hardest ticket types; run a short practice hour before the September hires start. Stopping at the first idea would have missed two of them.

On video, give the person room to think. Silence feels longer on a call than it is, and the lead's reflex is to rescue the pause with advice. Count to five before you speak. Don't judge options while they're being collected either: "that won't work" after the first idea stops the list. Sorting and choosing come at the next stop, and the person does it.

Asking more than you tell doesn't mean hiding what you know. A manager-coach also has context and experience the person lacks, and withholding a useful idea out of coaching purity helps nobody.

The rule is about timing and weight. Share your idea after the person has put their own options on the table, and offer it as one option among several, with a question attached: "One more idea: present the rota to Iryna yourself. What do you think?" Then let them weigh it like any other.

What spoils coaching is leading with your answer, before they've thought, or pushing it as the right one. Once the lead's idea is out first, the person's own options tend to shrink to variations of it, and the step they choose becomes yours again. If they've truly run out of ideas, offering yours is not a failure of coaching; it's part of it.

The conversation ends with one next step the person owns: a concrete action, a date and a way to report back, written in your shared one-to-one notes. Olha's: by the 20th she writes up the five hardest ticket types, and for two weeks she pairs instead of taking over; she brings both to the next one-to-one.

Her step, her words, her date. If you write it for her, it quietly becomes your plan, and her commitment drops with it. One step is enough; a list of six is a wish list.

And if the step doesn't happen? That's simply the next reality question, asked without blame: "What got in the way?" Maybe the August queue ate the time, maybe the step was too big. Either way, the answer leads to a smaller or re-dated step that's still hers. Leading with the broken promise invites excuses; quietly doing it yourself, or skipping it, tells her the step never mattered.

A coaching conversation follows a simple order, and each stop depends on the one before.

1. Goal: "What would you like to be able to do by December?" It gives the conversation a direction.
2. Reality: "What have you tried so far, and what got in the way?" It grounds the goal in facts.
3. Options: "What could you try? And what else?" Options only make sense once both the goal and the reality are clear; before that, they're guesses.
4. Next step: "What will you do first, by when, and how will I hear about it?" One step the person owns.

The order isn't a script. A good conversation sometimes loops back: the reality may show the goal is too big for December, and you return to it. But skipping a stop costs you: options without reality solve the wrong problem, and a conversation without a next step is a pleasant chat that changes nothing. In a 30-minute one-to-one, all four fit.

Coaching works when the person can find the answer. In three situations they can't, and questions only leave them guessing.

• Unclear expectations. If nobody ever said how detailed a handover note should be, asking for her opinion is a guessing game. Set the expectation first.
• A skill gap. A new coordinator who has never issued a refund in the billing system needs teaching and a worked example, not questions. People new to a task learn best from clear instruction; coach more as their competence grows.
• A performance problem. When a pattern keeps hurting families, like promised callbacks missed for the third month, say it directly: behavior, impact, next step, as in the lesson on feedback. Weeks of gentle questions only postpone a talk that is overdue.

Maryna's wish to explain the refund policy better to upset parents is a coaching case: she knows the policy well and her tickets reopen least on the team. The goal is hers and the answer is within her reach.

Sometimes a one-to-one turns from work to the person: weeks of bad sleep, dread before every shift, something at home, something heavier. Coaching is the wrong tool here too. A lead isn't trained to dig into health or personal crises, and probing questions can do harm.

What a lead can do: listen without rushing to fix; take it seriously, without brushing it off ("everyone's tired in August"); ease what you control at work, such as load, shift pattern or a day off; and point to professional help, for example a doctor or a counselling service. Keep what you hear private, and never turn it into a review note or a performance risk; that punishes honesty, and people stop telling you things.

How it fits together: coaching walks a person from their goal through reality and options to one step they own, with your idea offered last; teaching, clear expectations, direct feedback or professional help take over where coaching can't. Next comes the review season: facts before ratings.