Almost every newly minted coach knows the feeling: training ends, you get the paper — and then silence. A few first clients, weeks between them, and no feedback on whether you're actually getting better.
In training you had a cohort, a trainer, weekly peer coaching. Now it's you and a client who won't tell you "you should've given me more room there." They'll only tell you whether they book again.
The gap between training and practice
This is the most fragile stretch of a coaching career. Not because you can't coach — you can, you just finished. But a skill you don't use and get no feedback on quietly rusts. And you won't notice, because no one's telling you.
The usual advice is "coach as much as you can." Except early on you don't have the clients, and the ones you do have you can't afford to treat as a practice range. You need to train in between them.
Practice that tells you the truth
Practice.coach exists for exactly this gap. You train with an AI client that behaves like a real person — it has a story, holds on to it, doesn't answer like a textbook. Run a session any time, without risking a relationship with a real client.
And then you get feedback along the ICF competencies — not a pat on the back, but specifically where you led instead of following the client, where you skipped an emotion, where a question could have gone deeper. The same thing a good supervisor would give you. Whenever you need it.
The goal isn't to replace real clients. It's to not rust before you have enough of them.
And once you do, you upload a real session transcript and get the same score as your practice — so for the first time you see drills and live work on one scale.
The paper from your training says you started. What you do next decides whether you're a better coach a year from now, or just a year older.