Relationship Miracle Questions: A Practical Coaching Guide
Nour
Founder, Coachful
Relationship miracle questions invite a client to describe what they would notice if an important relationship became more workable. A follow-up considers what another person might observe the client doing differently. In coaching, this can turn “We need to collaborate better” into a chosen action, without promising cooperation or access to the other person's thoughts.
The miracle question comes from Solution-Focused Brief Therapy, developed by Steve de Shazer, Insoo Kim Berg and colleagues. Related ideas have been adapted for coaching, including work by Peter Szabó. Clinical origins do not turn an ordinary coaching conversation into treatment. This guide develops an original workplace example in which a client declines the word miracle, a colleague declines a meeting and a legitimate customer constraint changes the plan.
Use the future alongside an honest present
A preferred-future question asks what would be present instead of repeated conflict: a clear request, a known decision owner, a respectful boundary or a completed handoff. It does not make facts, acknowledgment or accountability unnecessary. Careful listening, mediation or a formal process may be the useful next route.
| Coaching task | Original question | Keep clear |
|---|---|---|
| Describe a preference | What would a workable exchange let you do? | The client chooses the ambition, including respectful distance. |
| Make it observable | What action could someone see or hear? | No prescribed eye contact, tone or body language. |
| Consider another perspective | What might the colleague notice, if anything? | A hypothesis, not mind-reading. |
| Choose a step | Which part is appropriate and within your influence? | The client cannot guarantee another person's response. |
| Review | What happened, and what matters now? | No promised improvement or fixed timeline. |
The SFBTA's 2025 third-edition treatment manual describes preferred-future and relationship questions in a clinical framework, including considering what significant others might observe. That is background for the broad idea, not evidence that this coaching exercise repairs a workplace relationship. The questions, case and worksheets below are original non-clinical illustrations, not the manual's client dialogues or a fixed protocol.
Three teaching tasks, with permission to change direction
You can organize the conversation around describing a preferred interaction, making a possible difference visible and finding a foothold in experience. These are teaching tasks, not formal stages everyone must complete. Some clients prefer an ordinary description to an imaginative thought experiment.
Describe: choose a specific setting rather than an abstract future in which everyone is happy. “I can raise a missing handoff check without accepting work outside my role” may be a useful preference even if the colleague remains annoyed.
Make it visible: describe words, actions or decisions. “I will send one example with a clear question” is observable. “They will finally respect me” predicts someone else's interpretation. Explore the first without promising the second.
Find a foothold: ask whether some part has happened before and what conditions supported it. An exception during a quiet week may depend on staffing or an unusually simple request. It does not prove that today's problem is minor or that the client only needs to try harder.
If there is no useful exception, accept that. Consider a smaller question, support, an appropriate boundary or another route. Do not repeatedly ask until the client supplies evidence for the coach's model. A relationship may become workable without becoming close, forgiving or warm.
Introduce the question without forcing optimism
- Hear the situation: ask what happened and what the client wants from this conversation.
- Offer a choice: check whether imagining a different interaction would help, using ordinary language if preferred.
- Describe a visible difference: explore a sentence, decision, boundary or routine the client could recognize.
- Keep the colleague tentative: accept that they might not notice, agree or respond.
- Choose and review a step: consider authority, receiving capacity and a contingency, then examine what actually happens.
The word miracle may feel useful, unrealistic, religiously loaded or distracting. None of those reactions makes the client resistant. Preserve their purpose rather than defending a word. If improvement feels impossible, ask what they need to understand before choosing anything.
| Starting point | Possible original question |
|---|---|
| A broad complaint | Which ordinary exchange would you most want to make workable? |
| A vague goal | What would be different in the handoff itself? |
| A modest improvement | What would make the next request a little clearer? |
| A boundary need | What limit could you state and appropriately maintain? |
| An impossible future | What needs attention before we consider a next step? |
A client may choose an exit, a documented concern or formal support. Maintaining the existing relationship on its present terms is not the assumed goal. Individual coaching with one person also does not make the coach a neutral mediator for everyone involved.
If the client says the colleague would not notice any difference, accept that possibility. Ask what the client would notice themselves, or whether the action would still serve a purpose without recognition. Sending a complete request may make the record clearer even when the colleague remains unimpressed. Do not invent a more perceptive or generous version of that person to keep the exercise moving. The imagined perspective is a prompt for thought, not testimony.
A preferred future can also be a firmer limit. A client might want to stop accepting approval requests outside their role or after an agreed cutoff. Explore the authority, existing service obligations and receiving route that support the limit. A personal preference cannot silently rewrite an on-call commitment or customer requirement. The next step may be clarifying those arrangements before sending a new boundary. If the other person dislikes a properly supported limit, friendliness is not the only measure of whether it was appropriate.
Case: a preferred handoff meets a real constraint
In this original fictional case, Sofia coaches Ethan about repeated product handoff problems with colleague Bea. Ethan wants a shared handoff checklist by 20 June 2031. Sofia has no mandate to contact Bea, and no private coaching information is sent to her. Earlier business work, service obligations and the historical handoff examples are outside the measured account below.
Sofia spends fifteen minutes on Monday 2 June preparing and sending a finite plan: a forty-five-minute session on 4 June, a private written return by 9 June at 17:00, a twenty-minute review on 19 June and a closing read on 20 June. Ethan spends ten minutes reading and accepting on 3 June. Sofia spends ten minutes reading that acceptance and preparing before their session.
Both attend the forty-five minutes on 4 June. Ethan declines the word miracle and chooses an ordinary next-meeting description. This original ten-turn exchange occurs within their conversation:
Ethan: I want the handoff discussion to stop becoming a blame exchange.
Sofia: What would make today's work useful without needing a perfect relationship?
Ethan: A way to ask for a checklist and keep my approval boundary.
Sofia: Describe the first visible difference in a more workable meeting.
Ethan: I would bring one example and ask which checks must precede acceptance.
Sofia: What might Bea notice about that opening?
Ethan: Perhaps that I want a process, not an admission of fault. She might not see it that way.
Sofia: Which part can you choose if she still disagrees?
Ethan: I can make the request and avoid approving a change outside my role.
Sofia: What would help you handle a different response?
Ethan recalls a clearer handoff during a quieter week. That earlier event is a possible source of questions, not evidence that the same conditions exist now. He also chooses an informal scale: four means he can name the issue but cannot reliably finish with the checks and authority clear; five would include one agreed example and a named route for late changes. These are his descriptions, not a standardized relationship assessment.
Correct a prediction before turning it into a plan
On 5 June Sofia spends ten minutes preparing and sending V1 only to Ethan. It says Bea will respect the deadline once he changes his opening. Ethan spends ten minutes reading and rejecting that prediction on 6 June. He has not agreed that his manner controls Bea's response.
Sofia spends fifteen minutes on 9 June reading the correction, revising the account and preparing and sending this V2 repair privately at 16:00, one hour before the promised return:
I predicted that Bea would respect the deadline if you changed your opening. That was not established, and I have removed it. You chose to bring one example, ask about required checks and keep your own approval boundary. Bea may disagree or identify information we do not yet have. Your action can be worthwhile without producing her cooperation. Please use the revised plan only if it still fits your purpose; I will not contact Bea or share this reflection with her.
Ethan spends ten minutes on 10 June reading V2 and preparing and sending an ordinary business request to Bea. He proposes twenty minutes on 12 June to examine one handoff, identify required checks and discuss a short checklist. The message does not disclose coaching, claim Bea's motives or demand an admission of fault.
Bea spends ten minutes on 11 June reading and sending this counterproposal:
I cannot take the twenty-minute meeting on 12 June because of existing customer work. I can use fifteen minutes on 13 June to review the example and send a written answer, then five minutes on 17 June to read your selected status and give a brief return. I cannot agree in advance to a rule that all late changes stop. Some requests involve mandatory service checks. Please keep that question open rather than treating my meeting decline as refusal to cooperate.
Ethan spends ten minutes on 12 June reading the counterproposal and sending acceptance of the written route. Bea spends five minutes reading it that day. The twenty-minute meeting does not occur. Her alternative has an actual accepted time allocation, rather than an invisible expectation that she help after her other work.
The other person's reply changes the facts
Bea uses fifteen minutes on 13 June to review the existing handoff and prepare and send this return:
The example leaves out a mandatory customer check that had not been completed when the late change arrived. I agree that our handoff needs a clearer record, but not that every late request can be rejected or that your deadline overrides the check. My concern is the service obligation, not your willingness to sound positive. Please correct the example before asking for a shared checklist. The authority for any delivery exception also needs to be established through the appropriate work process.
On 16 June Ethan spends fifteen minutes reading that return, correcting his example and six-field plan, and preparing and sending the following status to Bea. A warmer relationship is no longer his test of whether the action was useful:
I omitted the mandatory customer check from the example. I have corrected that omission and withdrawn the suggestion that a cutoff can override it. I still need a clear handoff record and will not approve an exception outside my role. A shared checklist would need to include the check, the relevant authority and a workable route for later requests. This is a proposal for clarification, not a new rule you have accepted or an instruction to perform extra work.
Bea's five-minute block on 17 June includes reading and sending a brief reply. She acknowledges the factual correction but does not agree to a general cutoff or new work. Ethan spends fifteen minutes on 18 June reading that reply and preparing and sending selected learning to Sofia. The receiving work is included; he does not claim that Bea's acknowledgment completes a shared checklist.
A filled six-field plan, with the boundary retained
| Field | Ethan's revised account |
|---|---|
| Situation | A handoff discussion involving a late request and a mandatory customer check. |
| Chosen action | Correct the example and ask what a usable checklist must include. |
| Purpose | Clarify required information and decision authority without demanding an admission of fault. |
| Limit | No approval outside Ethan's role; no assumption that a cutoff overrides the service check. |
| Contingency | If authority remains unclear, use the appropriate work route rather than keep negotiating a personal promise. |
| Review | Examine what was sent, received, corrected and still unagreed on 19 June. |
A blank version uses the same six fields. Ask who would need to receive or act on a proposal, whether they have accepted that work and what happens if they decline. A plan should not depend on the client becoming endlessly calm or accommodating. A firm and appropriate boundary can be part of a preferred future.
Sofia spends ten minutes on 19 June reading Ethan's learning; Ethan spends five preparing before their twenty-minute review. They retain the actual request, Bea's legitimate alternative, the omitted check and its correction. The shared-checklist target remains unmet. No delivery exception, escalation appointment or additional work has been authorized in this exchange.
Ethan prefers a written description to repeating his original scale after learning about the missing check. Sofia does not invent a comparable five, a half-point improvement or evidence of relationship repair. The initial four was a conversation aid with his own meaning, not a score that the method must improve.
Sofia spends ten minutes after the review preparing and sending this finite close. Ethan spends five minutes reading it on 20 June without replying:
You made the request, accepted a written alternative and corrected an omitted service check. Bea has not agreed to a general cutoff, and the shared checklist is unfinished. Keep your approval boundary and the real customer constraint visible. Your next work decision needs appropriate authority and accepted capacity; it is not created by this coaching note. We have no evidence that Bea now respects a deadline or that the relationship has been repaired. No further reply is due under this finite plan.
Count the attempted change and the receiving work
| Person | Measured activities, minutes | Used / allocated |
|---|---|---|
| Sofia | 2 June purpose and finite plan prepared/sent 15; 4 acceptance read and preparation 10; session 45; 5 V1 prepared/sent 10; 9 correction read, V2 revised and repair sent 15; 19 learning read 10; review 20; close prepared/sent 10 | 135 / 165 |
| Ethan | 3 plan read and accepted 10; 4 session 45; 6 V1 read and correction sent 10; 10 V2 read and business request prepared/sent 10; 12 counterproposal read and acceptance sent 10; 16 return read, example and plan corrected and status sent 15; 18 brief reply read and selected learning prepared/sent 15; 19 preparation 5; review 20; 20 close read 5 | 145 / 180 |
| Bea | 11 request read and written alternative prepared/sent 10; 12 acceptance read 5; 13 handoff reviewed and return prepared/sent 15; 17 status read and brief reply sent 5 | 35 / 50 |
The known total is 315 person-minutes against 395 allocated, a difference of eighty: thirty for Sofia, thirty-five for Ethan and fifteen for Bea. The two coaching conversations contribute 130 already included participant-minutes, leaving 185 other measured minutes. Earlier business events, materials, setup and ongoing service work are outside the account, not free. The difference is not established spare capacity across the practice.
When a future question is the wrong next move
If the client describes harassment, discrimination, threats, coercion or serious harm, do not use an imagined better relationship to move past facts and necessary support. Work within your role and help identify suitable organizational, professional or urgent routes. A positive description does not remove accountability.
Joint work needs its own agreement about role, confidentiality, participation and decisions. Individual coaching does not authorize contacting or characterizing an absent colleague as a second client. The ICF Code of Ethics supports clear role and information-exchange agreements; applicable requirements still need separate consideration.
Clinical relationship work or trauma treatment requires appropriate competence. The NIMH psychotherapy overview describes treatment under qualified mental-health guidance. A coaching adaptation is not a treatment merely because its ideas have therapeutic origins. Seek appropriate referral or collaboration where needed.
Review without making the client responsible for cooperation
Ask what they did, what response they observed and what changed in their understanding. A colleague may remain difficult or raise a legitimate constraint. A request can be useful without producing agreement; sometimes it reveals that the preferred plan was incomplete.
The next step may be a firmer boundary, a factual correction, appropriate facilitation, support or an exit. Do not promise that a gesture disarms a rival or that relationships normally improve within twenty-four to forty-eight hours. Review the actual context instead of assigning the client responsibility for someone else's behavior.
If a scale is useful, define the endpoints and observable meaning with the client. If meaning changes, a new issue appears or a written account fits better, do not force a numerical comparison. The goal is an appropriate next interaction, not a rising chart.
Organize reflection without implying permission to share
Coachful is our #1 recommendation for an appropriate preparation worksheet and agreed follow-up. That is an editorial workflow recommendation, not evidence of relationship outcomes. Forms and worksheets can hold the original six-field plan; Coach review is private to the coaching team. Saving it does not deliver the repair to Ethan or notify Bea.
Use the separate agreed return route, check access and keep private reflection distinct from a business request. A coach's own client-linked task does not assign work to the colleague. The case uses manual selected correspondence, not an automatic reminder, relationship assessment or customer-account trial.
For related practical distinctions, see solution-focused scaling and client confidentiality. Explore relevant Coachful features without treating supplied demonstration screens as this client's actual records.
Frequently asked questions
What is a relationship miracle question?
A thought experiment about a preferred relationship in observable terms, including what the client or someone else might notice. It is not a prediction or an objective assessment.
Do both people need to attend?
No. Individual coaching can explore the client's choices. Their account does not establish the absent person's perspective or appoint the coach mediator for both.
What if improvement feels impossible?
Acknowledge that and consider a smaller question, needed facts, a boundary or another support route. Do not repeat the prompt until the client supplies an optimistic answer.
Does it replace discussing past harm?
No. Facts, acknowledgment, accountability and formal processes may matter. Future questions should not bypass them.
How quickly should the relationship improve?
No timeframe is guaranteed. Review the chosen action, real response and context. Another person's disagreement is not proof that the client performed the exercise incorrectly.
Background: SFBTA's Treatment Manual for Working with Individuals, third edition (2025), distinguishes preferred-future and relationship questions within therapy. Peter Szabó and Insoo Kim Berg's Brief Coaching for Lasting Solutions (2005) provides historical coaching context; no full-book reading or formal protocol is claimed here.
