Session Observations That Show Development: Tracking Growth Across Sessions

Quick answer. From January 1, 2027, ICF mentor coaches document each observed session on a Session Observation Form and the cumulative picture on a Competency Review Form. ICF's own competency model requires mentor coaches to identify "trends in the mentor coaching client's behaviors and competence across sessions" — but no ICF form contains a field for a trend. The Session Observation Form is a per-session snapshot; the Competency Review Form is a binary, session-agnostic checklist. The cross-session record that makes development visible is the mentor coach's own to construct — and how much movement the forms can even register differs sharply between the ACC and PCC versions. Sources: ICF Mentor Coaching Competency Model, PCC Session Observation Form, PCC Competency Review Form

A question making the rounds among mentor coaches this month, paraphrased: observing a coaching session is not just listening and writing down whatever catches your attention — so how do you conduct session observations that truly show a coach's development? What behaviors are present, and where? Which competency do they demonstrate? What patterns repeat? And when you observe several sessions over an engagement: what was happening at the beginning, what has changed, which behaviors are becoming consistent?

The one-line version of that post's argument is worth keeping on your desk through 2027: one observation gives you a snapshot; a well-documented observation process gives you a developmental trajectory.

That is exactly the philosophy ICF has adopted — and, as we'll see, it is written into the competencies and the instructions more thoroughly than into the paperwork. This article is the practical companion to our guide on what the forms require you to write: not the writing, but the observing — how to design and document a multi-session observation arc so that development actually shows.

Important disclaimer. Mentor Coaching AI is not affiliated with or endorsed by the International Coaching Federation. We summarise publicly available ICF announcements, forms, and documents. Policies evolve — verify details directly at coachingfederation.org before making decisions.

Why one observation is a snapshot — in ICF's own words

The clearest statement of why the old model is being retired comes from the ICF Mentor Coach Handbook: coach assessors evaluate "a snapshot of the coach's skill level based upon a single coaching session," whereas mentor coaches work "through formative evaluation of the coach's skill demonstration over a period of time."

ICF's accreditation-side announcement makes the design intent explicit: "Rather than evaluating coaching ability at a single point in time, this approach allows learners to demonstrate competency across multiple interactions and learning experiences" (ICF Standards & Compliance Updates). And the Mentor Coach Specialization announcement describes the new evaluation as "multiple observations with feedback over time," with mentor coaches expected to "observe coaching sessions, calibrate performance against credential criteria, deliver actionable feedback, and document demonstrated skills across the engagement."

The requirement behind all of this, for ACC and PCC Portfolio applicants from April 1, 2027: "at least three of your coaching sessions must be observed and reviewed by your mentor coach" within the 10-hour, minimum-three-month mentor coaching engagement. Three observations, spread across at least twelve weeks, with a Session Observation Form (SOF) for each and a Competency Review Form (CRF) closing the engagement.

Three snapshots do not automatically make a trajectory, though. Whether they do depends on how you observe.

The trajectory requirement is in the competencies — not in the forms

Here is the asymmetry at the heart of the 2027 model, and every piece of it is quotable from an ICF document.

The competency model requires cross-session pattern recognition. Competency 4 of the ICF Mentor Coaching Competencies ("Conducts Formative Appraisals") includes behavior 4.5: "Identifies trends in the mentor coaching client's behaviors and competence across sessions to determine skill development." This is the only place in ICF's published framework where cross-session observation is named as a required mentor coach skill.

The Competency Review Form instructs you to think in sessions. Both the ACC and PCC CRFs open the skills section with: "Indicate in which session(s) the client demonstrated each required skill at a competent level."

But the answer fields are binary and session-agnostic. Under that instruction sit two checkboxes per skill: Proficiency Demonstrated / Proficiency Not Yet Demonstrated. There is no field in which to record which session, no field for a trend, no field for movement. The CRF's session log — date, modality, duration of each observed session — is ICF's only structured across-time artifact, and it records that sessions happened, not what developed between them.

The practical consequence: the cross-session record that satisfies Competency 4.5 — and that makes your Competency Review Form honest rather than impressionistic — lives outside ICF's paperwork, in whatever observation system you keep. The forms assume you have one. They don't give you one.

This isn't a design failure so much as a division of labor. The CRF enforces traceability — "any skills documented in the Competency Review Form must also be documented in the appropriate Session Observation Form" — so the per-session record is the evidence base. Turning three evidence bases into a developmental story is precisely the professional skill the Mentor Coach Specialization certifies. But it means the working mentor coach needs a method, not just the forms.

ACC and PCC forms measure differently — and it changes what development can show

An under-reported fact about the June 2026 form releases: the ACC and PCC Session Observation Forms do not use the same rating scale.

The PCC SOF is binary. Each PCC Marker under Competencies 3–8 is rated Observed / Not Observed, with an "Evidence of Marker Demonstrated" field. The PCC Markers remain the behavioral reference — the form's instruction is "evaluate the session using the PCC markers below."

The ACC SOF is graded. Each behavioral statement is rated on a four-point scale plus N/A: Exceeds Standard / Meets Standard / Below Standard / Does Not Meet Standard, where "Below the standard" is defined as partial demonstration. The form points to the ACC BARS (behaviorally anchored rating scale) resource guide for calibration.

Follow the consequences across three observed sessions:

• At ACC, the forms themselves can register a trajectory. A skill moving Below Standard → Meets Standard → Exceeds Standard across sessions one to three is development, documented in ICF's own instrument. Note the counting rule, though: on the ACC CRF, only Exceeds or Meets count toward "Proficiency Demonstrated." A skill rated Below Standard in all three sessions — partially demonstrated every time — is still Not Yet Demonstrated. Partial credit accumulates into feedback, not into proficiency.
• At PCC, the forms can only register presence. Not Observed → Not Observed → Observed is the closest the paperwork comes to showing growth — and the CRF marks a skill proficient if it was Observed in any one session observation form. Binary ratings cannot distinguish a marker that appeared once, shakily, from one that showed up in every session with increasing depth. If you want that distinction — and formative feedback needs it — your evidence notes have to carry it, because the checkboxes cannot.

Either way, the free-text evidence fields do the developmental work. The checkboxes decide the credential outcome; the evidence record is where the coach's growth becomes legible.

What observable behaviors are present — and where do they occur?

The per-session discipline that makes cross-session comparison possible later is boringly consistent: evidence, located.

Both SOFs carry the identical instruction: "Please list evidence from the session, such as timestamps or quotes, to support your selected rating option" — and if you observed a live, unrecorded session, to state explicitly that "no timestamps were recorded in the session." The feedback block then requires "a minimum of 2-3 areas of strength and 2-3 areas for development, accompanied by evidence (timestamps from a transcript, quotes from the coaching conversation, or what the coach may have said in a live session)."

Three practices turn this compliance requirement into trajectory raw material:

1. Keep the evidence format identical across sessions. A timestamp plus a short verbatim quote plus the competency it demonstrates, in the same structure every time. Session-three evidence you can lay side-by-side with session-one evidence is comparable; free-form impressions are not. This is also what ICF's timing guidance supports — reviewing the session "within two business days" while the observation is fresh.
2. Log what you didn't observe, with context. At PCC especially, "Not Observed" carries no information about why. A marker that had no natural opening in this client conversation is different from one the coach visibly avoided. One sentence of context per unobserved marker is the difference between a gap and a data point.
3. Note the conditions. Client, session type, what the coach said they were working on. ICF's handbook envisions the coach practising "across several different coaching sessions with several different clients" between mentor sessions — so your three observations will likely involve different clients and different presenting topics. Comparing across them honestly requires knowing what varied.

The unglamorous footnote: ICF is explicit that "time spent reviewing the recording and/or taking notes does not count towards the mentor coaching hours." The observation discipline described here happens on the mentor coach's own time — which is exactly why a repeatable method matters.

What patterns repeat? Building the cross-session layer

Between the SOFs (per-session) and the CRF (end-of-engagement binary), the missing middle layer is a cross-session tracker — the document that answers Competency 4.5. Practitioners converging on the new model describe essentially the same structure, whatever tool they build it in:

• One row per competency or marker, one column per observed session, populated from the SOF ratings and evidence. At ACC the ratings themselves show movement; at PCC, use your evidence notes to grade what the checkbox flattens (e.g., observed once, prompted vs. observed throughout, unprompted).
• A "carry-forward" list: the 2-3 development areas from the last SOF become explicit observation targets for the next session. This is the single highest-leverage habit. The SOF requires you to name 2-3 development areas each session; deliberately re-observing those same areas next time is what makes progress demonstrable rather than anecdotal. If session one's development area was "coach fills silences" and session three's evidence shows three held silences with timestamps, you have a documented trend — exactly what the cumulative feedback section of the CRF asks for ("clear, evidence-based feedback... for 2-3 competencies for both strengths and opportunities for development").
• The client's own reflections as parallel data. Each SOF includes a client pre-review and post-review reflection, the latter naming "2-3 skills for your continued development" — and the CRF's closing reflection asks the client to "summarize your progress... throughout the mentor coaching engagement." Where the coach's self-assessment converges with (or diverges from) your observation record across sessions is itself formative material.

One warning from ICF's own handbook while building any marker-grid tracker: the minimum skills requirements "should not be used... as a checklist or formula for coaching," and "the ICF Core Competencies should be the primary focus of the Mentor Coaching feedback." The tracker is your instrument, not your feedback. Patterns in it inform the conversation; they don't script it.

How do you compare the beginning with the end? Designing the arc

Because the engagement spans at least three months — ICF specifies "12 full weeks must pass between the first and final mentor coaching sessions," with a recommended two weeks between sessions for the reflection-and-practice cycle — the mentor coach has real design freedom in when the three-plus observations fall. Some verified anchor points for planning the arc:

• There is no recency requirement for the observed client sessions, so an early observation can happen almost immediately — the form itself states "there is no recency requirement for client coaching sessions." An early baseline observation maximizes the window in which development can occur and be re-observed. (This is one of several ways the observation model is more forgiving than the old recording-and-transcript submission it replaces.)
• Observed sessions must be real coaching. The handbook distinguishes mentor coaches from program observers precisely on this point: "mentor coaches observe coaching between a coach and an actual client," with the client's written permission for recording or live observation.
• The closing session is explicitly retrospective. The handbook defines its purpose: review "the client's cumulative demonstration of coaching skills across all observed coaching sessions" through the CRF, and "reflect upon and acknowledge the progress and growth the MC client accomplished over the course of the engagement." If your observation record can't support that conversation — if it shows three disconnected snapshots — the closing session exposes it.
• The record follows the coach. Clients who work with more than one mentor coach at a level "should provide future mentor coaches with the completed competency review form from prior mentor coaching engagements." Your documentation is the next mentor coach's baseline; write it so a stranger could pick up the trajectory.

A reasonable default arc for a three-observation engagement: baseline observation in weeks 1–2, second observation around the midpoint after two feedback cycles, final observation in the last fortnight — with the carry-forward development areas linking each to the next. ICF doesn't prescribe this spacing; the twelve-week minimum and the Kolb-style practice cycle it cites simply make it the natural shape.

Where tooling fits — honestly

Everything above compounds into one workload statement: per observed session, on unpaid time, produce timestamped, competency-tagged, consistently formatted evidence — then maintain a cross-session layer ICF's forms don't provide. That layer is where Mentor Coaching AI sits. It analyses each session recording or transcript against the ICF Core Competencies and produces timestamped, competency-mapped evidence in the same structure every session — which is precisely the property that makes session one comparable with session three. The judgment about what a pattern means, the feedback conversation, and every signature on every form remain the mentor coach's; what the tooling removes is the re-listening and manual evidence assembly that make longitudinal observation expensive. A free competency analysis of one session shows what the per-session evidence layer looks like.

The test for any observation system — spreadsheet, notebook, or software — is the question from that practitioner post: do your session observations make the developmental trajectory visible? If a colleague could read your records and tell you what changed between the first observation and the last, they do. If they'd see three unrelated snapshots, the 2027 forms will be filled in and the actual mandate — formative development over time — will be missed.

FAQ

How many coaching sessions must be observed?

ICF states that "at least three of your coaching sessions must be observed and reviewed by your mentor coach" for ACC and PCC Portfolio candidates under the new model. ICF has not published different minimum counts for ACC versus PCC. A Session Observation Form is completed for each observed session.

Can the observations be live, or must sessions be recorded?

Both. The Session Observation Forms state sessions "may be observed live or through an audio or video recordings," and the CRF session log offers in-person, virtual, and recorded as modalities. For live unrecorded sessions, the form requires noting that no timestamps exist — quotes and described moments become the evidence.

Do the observed sessions have to be with real clients?

Yes. The handbook specifies that mentor coaches observe "coaching between a coach and an actual client," with the client's written permission — unlike program-internal observation of practice between learners. (On consent mechanics, see our recording consent and ethics guide.)

Is there a form field for recording development trends across sessions?

No. The Competency Review Form instructs the mentor coach to "indicate in which session(s)" each skill was demonstrated, but provides only a binary Proficiency Demonstrated / Not Yet Demonstrated checkbox per skill. Cross-session trend records — required by Mentor Coaching Competency 4.5 — are the mentor coach's own documentation to maintain.

Does a skill count as proficient if it only appeared in one session?

Yes, by ICF's stated counting rule — with a level difference. At PCC, a marker marked "Observed" in any one Session Observation Form counts as Proficiency Demonstrated. At ACC, only sessions rated "Exceeds Standard" or "Meets Standard" count; "Below Standard" (partial demonstration) never converts to proficiency, no matter how often it appears.

Are these forms required for credential renewals?

No. ICF's handbook states the Session Observation and Competency Review Forms "are not required for a credential renewal application but can be useful tools for development." See our renewal guide for what 2027 changes for renewals.

Sources

• ICF — Mentor Coaching Requirement Replacing ACC & PCC Performance Evaluations
• ICF — Introducing the Mentor Coach Specialization
• ICF — Standards & Compliance Updates
• ICF Mentor Coach Handbook (PDF)
• ICF Mentor Coaching Competency Model (PDF)
• ICF — ACC Session Observation Form
• ICF — PCC Session Observation Form
• ICF — ACC Competency Review Form
• ICF — PCC Competency Review Form

If you mentor coaches, try the exercise before 2027 makes it mandatory: take two recordings of the same coach, a few weeks apart, and write evidence notes for both in an identical format. Then answer, from the notes alone, what changed. That answer — not either form — is the skill the new model is actually asking for.