When Your Observed Coaching Session Goes Badly: Recovering from Tough Mentor Coaching Feedback

Quick answer. One rough observed session is a data point, not a verdict. The pattern behind most of them is the same: heavy preparation, a list of questions on the desk, and a coach so busy managing their own performance that they stop listening to the person in front of them. The fix is not more preparation — it's shifting your attention from "what do I ask next?" to "what am I noticing?" And the structural news is good: from April 1, 2027, ICF retires the one-shot performance evaluation for ACC and PCC Portfolio candidates in favor of a formative model where your mentor coach observes at least three sessions over time. The era of one recording deciding everything is ending.

A coach in the final stretch of an ICF-accredited program recently described an experience in a coaching forum that hundreds of coaches will recognize. An observed peer session with a supervisor watching. Nerves. Heavy preparation, including a bank of questions within reach. And then the session that never went anywhere — surface-level, practical, safe. The feedback afterward: you didn't get under the surface, and several competencies weren't demonstrated. The coach's own words captured the spiral that followed: frustrated, confidence shaken, and a bigger group observation the next day they were dreading.

The replies from working coaches were strikingly consistent — and they map almost perfectly onto what the ICF Core Competencies actually ask for. This article works through that overlap: why observed sessions go wrong in this particular way, what "under the surface" means in competency terms, how to recover before your next observation, and why the 2027 changes to ACC and PCC evaluation are quietly designed around exactly this problem.

"I got in my head massively" — why observed sessions feel so different

Coaching is an attention profession. A session works when your attention is on the client — their words, their energy, what they're not saying. An observer in the room (or on the call) gives your attention somewhere else to go: to the evaluator, to the competency checklist, to the running commentary in your own head about how it's going.

That split attention isn't a character flaw; it's the predictable cost of being assessed. But it has a specific, visible effect on the session. The 2025 ICF Core Competencies define Listens Actively (Competency 6) as focusing "on what the client is and is not saying to fully understand what is being communicated." You cannot do that while simultaneously monitoring yourself. Something has to give, and under observation, it's usually the listening.

Two things follow from this. First, a session that goes flat under observation is weak evidence about your coaching — it's mostly evidence about your nervous system under assessment. Second, the way out is not to try harder at both jobs at once. It's to deliberately hand the self-monitoring job to someone (or something) else — which is precisely what mentor coaching, recordings, and post-session review are for.

The question-bank trap

The most common preparation strategy before an observed session is also the most reliably counterproductive one: script your questions in advance.

It feels safe. It is the opposite. A prepared question list anchors the session to the conversation you imagined yesterday, not the one happening now. Experienced mentor coaches spot it instantly, because the tell is structural: the coach's questions stop connecting to what the client just said. The 2025 competency language is explicit that customization is the skill — Listens Actively asks the coach to consider "the client's context, identity, environment, experiences, values and beliefs" to understand what the client is communicating, and Evokes Awareness (Competency 7) asks the coach to consider "client experience when deciding what might be most useful" and to adjust "the coaching approach in response to the client's needs." A list on the desk cannot do any of that.

The practitioners replying to that forum thread said the same thing in plainer words: understand why your prepared questions were good questions — open, responsive to shifts in tone or posture — and then generate questions with those properties live, from what you're hearing. One suggested keeping a loose mental map of the session arc (agreement, exploration, insight, forward movement) instead of a script: waypoints, not a route.

If you need a bridge while you build that trust, make it as small as possible. A card with two or three deepening prompts — "tell me more," "what else?", "what's unsaid so far?" — buys you thinking time and opens the client up, without dragging the session back to yesterday's plan.

What does "getting under the surface" actually mean?

"You stayed at the surface" is among the most common — and most confusing — pieces of observed-session feedback, because it sounds like a personality critique. It isn't. In competency terms it points at specific, learnable behaviors:

• Noticing that there's more. Listens Actively includes recognizing and inquiring "when there is more to what the client is communicating," and noticing and exploring "the client's emotions, energy shifts, non-verbal cues or other behaviors." Surface-level sessions are usually full of these signals — the coach was too busy to catch them.
• Coaching the person, not the problem. Evokes Awareness is about facilitating insight — questions that help the client explore beyond their current thinking about their situation and about themselves. A session that "mostly spoke about practicalities," as the coach in the thread put it, is a session that stayed entirely on the problem's logistics and never touched who the client is in relation to it.
• Letting silence work. Powerful questioning, silence, and metaphor are named tools of Competency 7. Nervous coaches fill pauses; the client's processing time gets crowded out by the next question. If you remember one mechanical fix, make it this: after a real question, stop talking.

Notice that none of these require brilliance. They require available attention — which is exactly what the question bank and the self-monitoring were consuming.

The 48 hours after tough feedback

What you do immediately after a stinging observation matters more than the observation itself.

Separate the session from the identity. "This session didn't demonstrate several competencies" and "I'm not cut out for this" are different sentences. Mentor coaching feedback is calibrated against level-specific behavioral criteria — that's the entire design. A below-the-bar session locates you on a development path; it doesn't remove you from it.

Extract the mechanism, not the mood. The coach in that thread had already done the most valuable part before anyone replied: they could name what happened ("I was relying on the questions in front of me rather than listening"). That's a diagnosis, and diagnoses are actionable. Write yours down in one sentence while it's fresh.

Get back in front of an observer quickly. Nearly every experienced voice in the thread gave the same advice: volunteer for the next observed session, not to prove anything, but because avoidance consolidates the fear and reps dissolve it. The coach's own update ended exactly there — hand up, pressure off.

Review the tape if it exists. Several coaches noted that we are unreliable narrators of our own sessions — one described a session she'd privately written off as her worst, only for the client to email about how helpful it had been. A recording or transcript replaces the story you're telling yourself with what actually happened. This is also where a session-analysis tool like Mentor Coaching AI earns its place in the loop: upload the session and get a competency-by-competency read — where you customized to the client and where you ran a script, where you inquired into an energy shift and where you drove past one — before your next observed session, in private, with no one watching.

Why 2027 makes one bad session matter less

Here's the structural context that the coach in that forum thread — and every candidate currently dreading a recording — should know.

Today, ACC Portfolio candidates submit one recorded session with transcript, and PCC Portfolio candidates submit two, for performance evaluation. The recording is a snapshot: one hour to demonstrate everything, with the exact "being watched" pressure described above baked into the format.

That model is being retired. Per ICF's announcement, effective April 1, 2027, the performance evaluation is no longer a requirement for ACC and PCC Portfolio credentials. In its place, candidates complete an enhanced mentor coaching requirement in which:

• At least three of your coaching sessions are observed and reviewed by your mentor coach — a mentor who, from January 1, 2027, must hold the new Mentor Coach Specialization (MCS);
• your mentor completes a Session Observation Form for each reviewed session, which you keep on file (and submit if audited);
• your mentor completes a cumulative Competency Review Form that goes into your application, consolidating observations across sessions to determine whether you've demonstrated the competencies at your level.

ICF explicitly calls this a formative evaluation model: observing multiple sessions, providing "actionable feedback that advances skill development session by session," and documenting growth over the engagement. The familiar scaffolding — 10 hours of mentor coaching over a minimum of three months, at least 3 of them one-to-one — stays in place; the observation and documentation layer is what's new. (MCC candidates still do a performance evaluation, though their mentor coaching hours also require an MCS mentor from January 2027.)

For a coach who freezes under one-shot assessment, this is a genuinely better deal: a below-par session becomes one observation among several, followed by targeted feedback and another chance to show the skill — the same "get another rep" logic the forum coaches recommended, now built into the credentialing process itself. It also raises the value of every observed session, because your mentor coach is now writing evidence down each time; walking in with a pattern already half-fixed — because you reviewed your own recordings between sessions — compounds fast. What your mentor coach writes on those forms, and how the formative model works from their side of the table, is covered in our guide to formative feedback in mentor coaching.

FAQ

My observed session went badly — will it count against my credential?

Under the current (pre-April 2027) model, practice observations inside your training program are developmental; what's evaluated is the recording you choose to submit — see our performance evaluation preparation guide. Under the 2027 formative model, your MCS mentor coach documents at least three observed sessions and completes a cumulative Competency Review Form across the engagement — so a single rough session is context, not a verdict, and improvement across sessions is exactly what the model is built to capture.

Should I stop preparing for observed sessions entirely?

No — but change what you prepare. Preparing content (scripted questions) works against you. Preparing state works for you: re-read the competency behaviors for your level, pick one focus (for many coaches: "pause after every question"), and arrive early enough to settle. Then let the session be about the client.

How do I practice "getting under the surface" between observed sessions?

Record your practice sessions (with consent — see our guide on recording and client consent), then review them against the competencies: find three moments where the client signaled "there's more here" — an emotion, an energy shift, a loaded word — and note what you did. Most coaches discover they heard the signal and drove past it. Naming that pattern is most of the cure; an AI competency analysis of the transcript can surface those moments for you when reading your own transcript cold feels impossible.

Does tough feedback mean I have the wrong mentor coach?

Not by itself — specific, evidence-based feedback tied to competency behaviors is the job, and from 2027 it's formally documented on ICF's standardized forms. But feedback should name moments and behaviors, not verdicts about you. If it consistently doesn't, or the relationship lacks the safety to bring your clunkiest sessions, that's worth addressing — our guide on finding the right mentor coach covers what to look for.

Sources

• ICF: Mentor Coaching Requirement Replacing ACC & PCC Performance Evaluations
• ICF: Introducing the Mentor Coach Specialization
• ICF: Performance Evaluations
• ICF: Mentoring & Supervision Requirements
• ICF Core Competencies (2025)
• ICF: PCC Competency Review Form

Mentor Coaching AI is not affiliated with or endorsed by the International Coaching Federation. Requirements change — verify details against coachingfederation.org before relying on them.