When I read that ICF was removing the final performance evaluation from major ACC and PCC credentialing pathways, my first question was not whether formative assessment could work.
Of course it can.
My question was simpler.
Who says no?
Beginning in 2027, the person helping a coach develop may also become the person who verifies that coach's competence. That could be a better reflection of how coaches actually grow. It could also replace one imperfect checkpoint with a different kind of vulnerability.
This is not a small administrative change. It moves the center of credential-level judgment from a final performance evaluated at one point in time toward evidence collected inside a developmental relationship.
I think there is a credible case for the change. I also think anyone who cares about the value of an ICF credential should be willing to examine what the new model asks us to trust.
What ICF is changing
First, some precision. ICF is not eliminating every performance evaluation for every credential.
Under the current Portfolio process, an ACC candidate submits one recorded session and transcript. A PCC candidate submits two. Trained ICF assessors review those recordings against credential-level criteria.
Effective April 1, 2027, the separate performance evaluation will no longer be required for new ACC and PCC Portfolio applications. Instead, mentor coaching delivered by a coach holding the appropriate Mentor Coach Specialization will become the primary method for validating competence. The mentor coach will observe coaching over time, document evidence through session observation forms, and complete a competency review.
ICF-accredited Level 1 and Level 2 programs are changing too. Beginning April 1, 2027, they will no longer be required to include a standalone Final Performance Evaluation. Programs must demonstrate how learners develop and demonstrate competence through ongoing formative assessment. They may retain a final evaluation, but it cannot replace evidence gathered across the learning journey.
The MCC performance evaluation remains. The written exams also remain. The ACC Exam still assesses coaching knowledge, while the PCC and MCC Credentialing Exam still evaluates situational judgment. This policy changes how actual coaching performance is validated for ACC and PCC. It does not determine whether candidates must understand and apply ICF standards.
ICF explains the transition in its ACC and PCC performance evaluation announcement and its education standards update.
What the current recording gets right
The existing model creates distance between development and judgment.
Your trainer can believe in you. Your mentor coach can know how hard you have worked. The assessor receives a recording and transcript, applies the standard, and decides whether the coaching demonstrated the minimum skills for that credential.
That separation matters. The assessor is not responsible for your confidence, your timeline, or the money you spent getting there. They do not have to preserve a working relationship after saying that the recording did not pass.
The process also creates a common checkpoint. Different candidates submit different conversations, but trained assessors use the same credential-level criteria. The session either contains sufficient evidence or it does not.
That does not make the judgment perfectly objective. Coaching is contextual, and assessment always involves interpretation. But an independent summative review provides something valuable: a person outside the development process has to agree that the evidence meets the standard.
What one submitted recording gets wrong
Here is the strongest argument for changing it: one selected recording is not the same as a coach's normal practice.
Candidates do not pull a random session from a hat. They record several, review them, ask mentors for feedback, and submit the conversation most likely to demonstrate the markers. ICF's own preparation guidance tells candidates to choose a session that showcases their skills.
That is understandable. It is also a narrow sample.
A coach may pass with one excellent session while demonstrating weaker habits the rest of the month. Another may coach consistently well and submit the wrong conversation. A client may bring a topic that does not create visible opportunities for every skill an assessor needs to hear. Audio quality, transcription, language, disability, and cultural communication patterns can add friction that has little to do with coaching competence.
PCC Portfolio candidates submit two recordings, which broadens the evidence. It still captures selected moments rather than development and consistency over time.
The current process is independent, but independence does not automatically make the sample representative.
What the new model could improve
The new model treats competence as a pattern rather than a performance.
A qualified mentor coach can observe several sessions, identify recurring habits, provide feedback, and watch what the coach does with it. The evidence can show whether the coach consistently establishes agreements, follows the client's language, notices shifts, and resists taking over, rather than simply whether those behaviors appeared in one strong recording.
That matters because coach development is not linear. A candidate may understand feedback immediately and need weeks to embody it. Another may demonstrate a skill once and lose it under pressure. Formative assessment makes that variation visible.
It can also improve access. Observation in the candidate's language and cultural context may preserve tone and meaning that a translated transcript loses. Candidates who use sign language or other communication methods may face fewer barriers. The administrative burden of creating transcripts and waiting for a separate assessment can be reduced.
ICF's new competency review forms are designed to consolidate evidence from multiple observations. On paper, that is richer evidence than one polished session.
The new model is not automatically less rigorous. It could be more valid.
The uncomfortable conflict inside the new model
The mentor coach is paid to help the candidate grow. The mentor coach may now also verify whether the candidate has demonstrated proficiency.
Those responsibilities can support each other. They can also collide.
What happens when the engagement is nearly complete and the evidence is not there? Does the mentor coach withhold verification? Recommend more paid sessions? Sign because the candidate has improved, even if improvement has not reached the standard? How does the candidate challenge a judgment made by the person with whom they have spent months building trust?
There is pressure on the other side too. A mentor coach may become overly conservative to protect their reputation or avoid scrutiny. Candidates could receive different thresholds depending on who they hire. The developmental relationship may become less candid if every observed mistake feels like evidence in an application file.
The weakness has moved. The old model risks overvaluing one selected performance. The new model risks relationship bias, inconsistent thresholds, and blurred financial incentives.
That does not mean mentor coaches cannot evaluate fairly. I mentor coaches, and I believe repeated observation can produce meaningful evidence. It means good intentions are not a control.
The standard will be as strong as "not yet"
The success of this change will not be measured by how many candidates find the process more supportive. It will be measured by whether the process distinguishes development from demonstrated readiness.
For that to happen, four things matter.
Mentor coaches need real calibration
Holding the specialization must mean more than completing training. Mentor coaches need continued calibration against shared examples, especially around borderline evidence. If two qualified mentors hear the same coaching and apply substantially different thresholds, the form does not create consistency.
Evidence has to be specific
"The candidate improved" is not the same as "the candidate demonstrated proficiency." Observation records should identify what happened in the coaching, which competency it supports, and whether the behavior appeared reliably across contexts.
Candidates need a clear path when readiness is disputed
A credible system needs transparency about what happens when the mentor coach cannot verify proficiency or when the candidate disagrees. More coaching may be useful, but purchasing additional time from the same evaluator cannot be the only imaginable answer.
Mentor coaches must be willing to say not yet
This is the heart of it. A mentor coach can be supportive, believe in the candidate, honor their growth, and still conclude that the evidence is not sufficient today.
If every completed engagement ends in verification, the assessment has become paperwork.
So, is removing the final evaluation a mistake?
Not necessarily.
A single high-stakes recording is a limited way to judge a relational practice. Evidence collected across multiple real sessions can tell us more about how a coach listens, partners, adapts, and learns. Moving assessment closer to development may produce stronger coaches instead of better audition tapes.
But the change is only stronger if the new model preserves the function the independent review served. Someone still has to apply the threshold. Someone has to document why the evidence is sufficient. Someone has to say not yet when it is not.
ICF is not removing the need to demonstrate competence. It is changing who gathers the evidence, when they gather it, and within what relationship.
That can work.
Now the profession has to prove that it will.
