Immunity to Change and Perfectionism: A Practical Guide
Nour
Founder, Coachful
Immunity to Change and perfectionism can be explored together when a client wants to change a specific pattern of extra checking, reassurance or revision. Start with the work's actual requirements. The useful question is whether a protective rule makes additional work feel compulsory after those requirements have been met.
Accuracy, care and appropriate review should survive the conversation. This guide explains the framework, distinguishes its main concepts and follows a complete fictional draft-review test. The outcome includes a real clarification, an accepted revision and continuing uncertainty. It does not establish a cure for perfectionism or a permanent developmental change.
Coachful is our #1 recommendation for organizing the coaching record, a chosen test, private coach preparation and an actual returned response. This is our editorial recommendation for that workflow, without a market-ranking or psychological-treatment claim. The framework's authors and the work's actual decision owners retain their distinct roles.
Understand what Immunity to Change proposes
Robert Kegan and Lisa Laskow Lahey developed Immunity to Change. Harvard Graduate School of Education's account describes examining competing commitments and limiting assumptions alongside a desired improvement. Its immune-system language is a developmental metaphor, not a demonstrated biological explanation of repeated revision.
Their 2015 HGSE interview distinguishes acquiring skills from changing aspects of mindset. Use that distinction carefully. A missing skill, an unavailable resource or a contradictory instruction still needs attention. A coach should not assume that every delay reveals a hidden psychological obstacle.
The authors' 2009 book, in authorized HarvardX excerpts, connects an improvement commitment, behavior working against it, protective competing commitments and underlying assumptions. Its test guidance emphasizes a safe, modest action, possible soon, undertaken to learn about an assumption. The book also cautions that an assumption is not necessarily false and one experiment is unlikely to settle it.
These primary accounts establish attribution and describe the approach. They are not controlled evidence that this article's perfectionism application produces a particular coaching outcome. The questions, working note, boundaries and case below are original applications; we do not reproduce the book's maps, guide sheets, source cases or full training process.
Distinguish care, protection and the predicted consequence
Begin with a requirement someone can explain
Identify the purpose and stage of the work. An internal discussion draft can legitimately contain a clearly marked open question. A released instruction or consequential approval may require a different level of verification. The coach does not decide that every document should use the same standard.
Ask who can confirm what the recipient needs. If nobody has agreed the scope, format or authority, clarify those matters before interpreting further revision. An employee receiving incompatible instructions may be responding to an actual management problem.
Describe behavior rather than assign an identity
“I repeatedly rewrite the introduction after the content check” is examinable. “I am a perfectionist” is too broad to define a useful change. Ask when the extra work begins, what it changes and which promised action it delays.
Look at reassurance as work too. Asking three colleagues whether a draft looks acceptable consumes their attention and can produce conflicting preferences. That does not make seeking expertise wrong. The question is whether each request has a necessary, distinct purpose or serves an attempt to remove every possible doubt.
Keep worry and commitment distinct
A worry describes a feared experience. A possible protective commitment describes what the client organizes their behavior to prevent. Someone may worry about being seen as careless and respond by trying to ensure that no visible imperfection reaches a reader. That is a tentative account of protection, not an accusation of dishonesty or laziness.
Both the improvement and the protective aim can matter. The client may value timely collaboration and being a careful colleague. The work is to investigate whether the particular protective strategy serves both values, rather than demand that care be abandoned.
Give the assumption a consequence that can be examined
A belief about what would happen makes the protective strategy seem necessary. Distinguish “a reader might request a revision” from “any requested revision means the whole contribution will be dismissed.” The first may be ordinary collaboration; the second predicts something broader.
Do not decide the conclusion in advance. If a workplace genuinely punishes questions or uses inconsistent criteria, that environment matters. If the client cannot identify a suitable, bounded situation, pause the experiment. No one owes the coach exposure to humiliation or professional risk as proof of commitment.
Preserve ownership and professional scope
The coach offers a possible account, then listens for correction. The client can change it, decline the framework or choose practical support instead. Completion of an impressive map is not the service goal.
The ICF Code of Ethics, revised May 2026, addresses competence, agreements and discussing changes in professional support when appropriate. Keep the purpose and support needs under review. A coaching exploration is not diagnosis, treatment or a maturity ranking. Significant distress or clinical needs require suitable qualified support and attention to whether this work should continue.
Recognize a contrasting case with incompatible instructions
Consider a separate illustrative situation: one reviewer requires a working summary of no more than ninety words, while another requires at least 150 words in the same summary. The client repeatedly rewrites it and misses a deadline. That behavior resembles Ren's delay, but the stated requirements cannot both be met by one version.
Do not begin by asking which hidden commitment prevents the client from finishing. First identify who can settle the instruction. Useful next choices might be one agreed version, two separately scoped versions or an authorized deadline change. Each has a real resource or authority implication. A coach cannot grant the deadline change by describing it as a learning experiment.
Even after the instructions are clarified, some extra revision might still occur. That later pattern could support a different inquiry if the client wants it. The distinction is sequential: remove the actual contradiction, then examine any remaining client-owned question. Otherwise an apparently thoughtful model can turn an organizational failure into a demand that the individual tolerate uncertainty while producing an impossible document.
Design a learning test that keeps the quality gates
- Choose the precise extra behavior. Specify what will change and what will remain. Stopping repeated cosmetic reassurance after a content check is different from omitting that check.
- Confirm a suitable task and recipient. Establish the draft's stage, audience, required controls and the recipient's actual review agreement. Keep consequential release decisions outside the experiment.
- Record the prediction and useful evidence. Identify the response that would support or qualify the assumption. Silence, an ambiguous expression or an imagined opinion may leave the question unanswered.
- Carry out the agreed action. Record any changes from the plan. A test that silently removes the real quality conditions may not examine the intended question.
- Interpret and decide. Correct genuine defects and examine the broader prediction separately. Continue, narrow the assumption, change the workflow or stop. Do not require a reassuring result.
Limit preparation for the test itself. A growing set of review prompts can become another perfectionist task. Agree enough detail to protect people and make the learning interpretable, without treating exhaustive documentation as the condition for taking any action.
Follow Ren's complete fictional draft-review test
Ren designs internal learning discussions. His coach is Jo, and his colleague Amal can choose the structure of an upcoming internal planning discussion. All people, messages, documents, observations and figures are fictional. The measured work runs from 4 to 10 May 2062. No actual customers, participants or confidential business material are involved in the draft.
State the standard and the pattern
Ren wants to send a useful twenty-minute discussion outline by 7 May at noon Amsterdam time. His previous pattern, as he describes it, is to complete the content and then seek three reassurance reads while repeatedly changing headings. That earlier work is contextual self-report, not measured work in this example.
The agreed quality gates are specific: the outline must total twenty minutes, state the decision being explored, identify two candidate questions, label unsettled responsibilities and contain no confidential client material. It is a working discussion outline, not an issued training instruction. Amal controls the internal discussion structure; neither Ren nor Jo can authorize customer research or a published process through this test.
At their thirty-minute session on 4 May, Jo initially suggests that Ren might be avoiding difficult content work. Ren corrects him: the content is usually finished before the reassurance cycle begins. Jo withdraws that interpretation. They use five minutes on the task, ten on Ren's account, ten on the prediction and five on test boundaries.
Complete the client's own account
Ren's written working note says:
I want to share an adequate working outline in time for a colleague to use it. I complete the content, then keep changing the presentation and ask several people whether it is ready. When I consider sending it without those reassurance reads, I worry that a requested revision will expose me as careless. My possible protective commitment is to prevent any reader from finding something they would change. The assumption I want to examine is that, if Amal asks for a revision, she will dismiss the whole outline as unusable. This account fits this situation; it is not a finding about all my work.
The worry, protective rule and prediction are different parts of the account. Ren wants to preserve care. The test changes his attempt to remove every possible revision before another person sees the draft.
They agree that Ren will create the outline, perform the required content check, allow one cosmetic pass of no more than ten minutes and send it without three additional reassurance reads. He will not insert an error, conceal an uncertainty or ask Amal to give him favorable feedback.
Obtain the actual receiving agreement
On 5 May Ren spends five minutes requesting a twelve-minute review by 7 May at 16:00. Amal spends five minutes reading and replying: “I cannot review by then. I can spend twelve minutes on it by 8 May at 10:00. I will check whether I can use the sequence and identify what needs clarification. I am not agreeing to approve a training release.”
Ren spends three minutes reading and accepting the counterproposal. Amal spends one minute reading his acceptance. Her later review still fits the 9 May noon internal planning deadline. Ren's own 7 May noon sending target remains unchanged.
The prediction's evidence is also agreed in advance. A specific request to change the outline, accompanied by an actual decision to retain useful parts, would qualify the predicted total dismissal. An explicit statement that the whole outline is unusable would support it for this instance, while still requiring attention to why. A polite receipt alone would not resolve the question.
Send the complete checked working outline
On 7 May Ren spends thirty-five minutes drafting, ten checking content and the time total, ten on the permitted cosmetic pass and two sending and verifying the version. He sends version one at 11:57, before his noon target. The work takes fifty-seven minutes within a sixty-minute allowance. He still feels uncertain and does not request another reassurance read.
The complete version-one outline is:
| Time | Working discussion outline, version one |
|---|---|
| 0–3 minutes | State the purpose: choose which onboarding-guide question to prototype next. This discussion does not authorize customer contact or release a new process. |
| 3–8 minutes | Compare two candidate questions: where do new starters lose track of the next action, and where do they lack a named contact? Identify what evidence would help choose. |
| 8–14 minutes | Draft one possible guide response for each question, using synthetic examples only. Mark evidence still needed. |
| 14–18 minutes | Choose a question to propose for a later prototype and identify the person who would own that prototype, subject to their agreement. |
| 18–20 minutes | Record the choice, uncertainties and required permissions. Open responsibility: the discussion recorder has not yet been agreed. |
The five intervals total twenty minutes. Both questions and the limited purpose are visible. The recorder is expressly unsettled; the document does not invent an agreement. No sensitive source material is included. All five declared quality gates have been checked before sending.
Respond to a real clarification
Amal uses her twelve minutes to read the outline and send feedback, returning it on 8 May at 09:40: “I can use the sequence and the two candidate questions. Please distinguish the recorder from the later prototype owner; those are different jobs. Would you be willing to record this internal discussion if I facilitate? I have not asked anyone to own a prototype yet.”
Ren initially experiences the requested revision as evidence of carelessness. He spends five minutes reading the actual message and separating that interpretation from what Amal said. She retained the sequence and questions. She also identified a real role clarification. Neither part should be discarded.
Ren uses eight minutes to revise and two to send version two at 11:00. He preserves the first four time intervals and replaces the final row with this complete text:
18–20 minutes: Ren records the chosen question, remaining uncertainties and permissions needed for the next step; Amal facilitates the discussion. The later prototype owner is a separate role and has not been agreed. Choosing a question does not authorize customer contact, commit a colleague to prototype work or approve a process release.
His covering answer explicitly accepts the recorder role for this internal discussion, not the separate prototype role. Amal spends six minutes reading version two, using its sequence in her internal planning record and replying at 11:30: “I have used version two's sequence and questions in my discussion plan. Your recorder role and my facilitation are clear. Prototype ownership remains open. This is the planning outline I will use, subject to any necessary arrangements before the discussion.” Ren spends three minutes reading it.
The actual use is an internal planning record before the 9 May noon deadline. The later discussion, participant agreements and prototype are not observed here. The case does not claim that a workshop was delivered or that a future project succeeded.
Interpret the test without abandoning standards
On 10 May Ren spends five minutes preparing for a twenty-minute coaching review. They use five minutes on actual events, ten on the assumption and five on his decision. Ren's update is: “Amal asked for a revision and used the outline. That does not fit my prediction that any requested revision means total dismissal. The role distinction was worth clarifying. I want to keep the content check and the one cosmetic pass, without treating three reassurance reads as a requirement for this kind of working outline.”
Jo writes a final response in Ren's individual client conversation: “Your evidence concerns this colleague and this working draft. A requested improvement coexisted with actual use. That qualifies the total-dismissal prediction; it does not show that every reviewer will respond similarly. Keep the quality gates and review the arrangement afresh for a different task.” Ren spends four minutes reading and keeping that note.
Ren has not agreed to an escalating exposure program or to apply the same rule to consequential releases. He still reports discomfort around requested changes. One completed test provides useful information without establishing permanent change, immunity removal or a new developmental stage.
Account for every contribution
| Person | Actual active minutes, 4–10 May | Total and allowance |
|---|---|---|
| Ren | First preparation 5; first coaching 30; review request 5; read and accept counterproposal 3; draft 35; content check 10; cosmetic pass 10; send and check version one 2; read feedback 5; revise 8; send version two 2; read actual use reply 3; follow-up preparation 5; follow-up coaching 20; read final coach note 4. | 147 of 155; 8 unallocated. |
| Jo | Initial preparation 10; first coaching 30; notes 5; follow-up preparation 10; follow-up coaching 20; write final returned note 5. | 80 of 85; 5 unallocated. |
| Amal | Read and answer request 5; read acceptance 1; review and return 12; read version two, use it in planning and reply 6. | 24 of 30; 6 unallocated. |
The whole test and coaching workflow uses 251 of 270 available person-minutes, leaving nineteen across separate people. Ren's fifty-seven-minute drafting block contains thirty-five minutes of drafting and twenty-two of checking, presentation and sending. His later revision and all receiving work are additional counted activity. Time for the future discussion or prototype has not been silently counted as completed work.
Keep the record and actual response clear in Coachful
- Prepare a client-owned worksheet. Invite the client to describe their task, observed extra behavior and tentative protective account in their own words. Keep the formal source attribution. Do not imply the software diagnoses hidden motives or administers an authorized proprietary assessment.
- Review the submitted account. Open the client's Reflections, then Forms & worksheets. The response's Coach review is private to the authorized coach team and excluded from the response PDF. That internal preparation is distinct from feedback delivered to the client.
- Return the promised feedback. Use the individual's profile message icon and client conversation with the agreed authorized audience. Jo's final note is sent there and Ren reads it. Keep actual workplace files within their approved organizational handling process.
- Record the actual chosen action. Link suitable tasks, habits or milestones to a goal. A task completion means the recorded action occurred; it does not prove an assumption false, remove discomfort or authorize future work.
Check information handling before collecting sensitive fears or workplace accounts. Unshared journal pages differ from pages a client explicitly shares through supported specialty workflows. Session recaps are automatic by default; approval can hold new unsent recaps, session-type overrides matter and changing a setting cannot recall a message already sent.
Review the relevant Coachful features alongside this workflow. The safe behavioral test guide develops test quality more broadly. Our intellectual-uncertainty guide examines a different question about asking a necessary clarification.
Adapt when the account or test does not fit
The standard is unclear. Clarify it with the authorized person. More self-reflection cannot settle which version a regulator, client or organizational owner actually requires.
The account feels imposed. Return to the client's words or leave the framework. Ren's correction of Jo's initial interpretation is essential information, not resistance to be overcome.
The result is critical. Examine the content and context. Correct a genuine problem. If there is ridicule, retaliation or inconsistent treatment, address the environment and appropriate support instead of treating it as a reason to intensify personal exposure.
The test changes midway. Record what actually happened and whether it still examines the intended prediction. Extra reassurance obtained secretly may change the question. A necessary safeguard added openly may reveal that another test is needed.
Mapping becomes another endless revision task. Agree a useful stopping point. A tentative account and one proportionate action may be sufficient; a perfectly completed worksheet is not evidence of progress.
Frequently asked questions
Is perfectionism always an Immunity to Change issue?
No. The client may face a skill gap, unclear requirements, conflicting demands or an unsafe environment. Investigate the actual work and their account before selecting a framework.
Does the test require deliberately making an error?
No. Ren retained the quality checks and changed repeated reassurance. His draft labeled an open responsibility honestly. Learning should not depend on creating avoidable risk for another person.
What if the assumption is partly accurate?
Retain the supported part and identify its scope. A particular reader may require a different format or react poorly. One event does not establish a universal rule in either direction.
Can a client value high standards and change this pattern?
Yes. The inquiry concerns the specific protective strategy and its consequences. Care, accuracy and appropriate review can remain while unnecessary reassurance changes.
What does one completed test establish?
Only what its actual design and observations support. Ren's draft was used alongside a requested clarification. That qualifies his prediction in this instance; it does not establish a treatment outcome or a permanent identity change.
