Before the change project: agree what you are taking on

Clarifying a change assignment means agreeing on the problem to investigate, the outcome sought, who can decide and what each party will contribute. It also means naming what you are not taking on. In consulting, this is often called contracting: establishing a workable agreement about the engagement. Here, the term describes a working relationship, not a substitute for a legal contract.

The conversation matters before an internal change lead accepts a brief just as much as before an external consultant submits a proposal. “Run a workshop to get people on board” sounds like a defined task. It leaves open whether the decision is still negotiable, what people are expected to do differently and whether the organization can support that behaviour.

Treat the first request as a starting point

Imagine a fictional wholesaler introducing an order-status dashboard. The service director asks for training because employees are still calling the warehouse. A facilitator could immediately design the session. A more useful first step is to ask what those calls accomplish and when they happen.

Perhaps employees do not understand the dashboard. Perhaps it displays information that arrives too late to answer customers. Perhaps people can read the status but cannot tell who owns an exception. Those possibilities point to different assignments. Training may remain useful, but its purpose should survive a look at the work.

GOV.UK’s discovery guidance explicitly recommends examining a prescribed solution and reframing it as a problem, including its boundaries and constraints. That guidance concerns public-service development; the transferable principle here is to clarify the need before committing to a particular intervention.

Ask for a recent example rather than a general diagnosis of the staff. What happened? What information was available? What did the employee do next? What consequence concerned the director? Keep “people resist the system” as an interpretation to examine, not an established explanation.

If there is not enough information, agree a limited discovery assignment. Its output might be a description of the handoff and a recommendation about the next step. That is an honest commitment; promising adoption before understanding the obstacle is not.

Find the person who can actually sponsor the work

The person who makes the request may not control the decisions needed to fulfil it. A service director might authorize training but not change warehouse staffing or software priorities. Naming them as sponsor does not create authority they lack.

The Association for Project Management’s account of sponsorship connects the role to governance, objectives and continuing accountability for benefits. It is more than lending a senior name to the launch. For this assignment, establish who can resolve the trade-offs that the work is likely to uncover.

Separate the sponsor, the day-to-day lead, the people doing the work and those affected by it. One person may hold more than one role, but the responsibilities still need to be understood. Ask who can approve a pilot, release staff time, change a process or stop an unsafe or unworkable activity.

Use stakeholder mapping to identify missing perspectives, not just powerful supporters. In the wholesaler, warehouse staff can explain how status information is produced; customer-service colleagues know which questions it fails to answer. A meeting attended only by the director and the dashboard supplier cannot stand in for either group.

Separate the outcome from the things you will produce

A workshop, a communication plan and a training guide are outputs. The desired outcome is what should become different in the work. Agree both, and be precise about what the engagement can reasonably influence.

In the fictional wholesaler, “deliver two training sessions” is an output. “Service employees can identify the next responsible person when an order is delayed” describes a proposed working outcome. “Make everyone enthusiastic about the dashboard” is neither an appropriate promise nor a clear operational result.

Agree how you will recognize progress. For an initial investigation, that might mean the team can trace several ordinary and exceptional orders and identify unresolved ownership. For a later pilot, it might mean observing whether employees can handle the agreed situations without inventing a workaround. These are suggested checks, not reported results or universal measures.

Include what must not deteriorate. Fewer calls would be a poor result if customers simply received confident but inaccurate answers. Faster handling would be questionable if unresolved cases became invisible. Discuss these counter-signals before presenting a single convenient target.

Where the baseline is unknown, make establishing it part of the assignment. Avoid retrofitting a success measure to whatever information happens to look favourable at the end.

Agree how the collaboration will work

The sponsor’s contribution belongs in the agreement alongside the facilitator’s deliverables. Who will make introductions? Who will protect time for observation? Who will resolve conflicting instructions? Who will explain a difficult decision to affected colleagues?

There is a useful parallel in the ICF’s 2025 coaching competencies. They call for explicit agreements about goals, responsibilities and the relationship, and for revisiting those agreements when necessary. Coaching and organizational consulting are different activities; the shared lesson is to make mutual expectations discussable rather than leaving them implicit.

Be equally clear about your own role. Facilitating a discussion is not the same as owning the business decision. Interviewing employees is not an agreement to evaluate their performance. Helping a manager explore options is not permission to promise those options on their behalf.

Discuss information handling before inviting people to speak. Explain what notes are taken, who can see them and how findings will be reported. Do not promise anonymity where roles or incidents make people identifiable. If the proposed reporting arrangement is incompatible with the conversation you want to hold, change the arrangement or the method before proceeding.

Also agree a practical route for disagreement. A useful working agreement allows someone to say that the brief no longer fits the evidence. It does not require them to manufacture a favourable answer because a workshop date has already been announced.

Put the essential decisions in a short record

Write down the agreement in language the participants would actually use. The following record is an original example for the fictional wholesaler, not a legal template or a prescribed project standard.

Agreement	Example
Immediate question	Why do service staff need additional calls when the dashboard shows an order status?
Initial scope	Examine the handoff between one service team and one warehouse
Output	A shared account of the handoff, unresolved questions and options for a bounded pilot
Sponsor contribution	Arrange access, release agreed staff time and bring process owners to the review
Change-lead contribution	Facilitate inquiry, distinguish observations from assumptions and document options
Exclusions	No system replacement, staffing decision or employee performance assessment
Decision point	Sponsor and relevant process owners choose whether to commission a pilot
Revisit trigger	Findings require another site, a different decision owner or substantially more staff time

Add named owners, actual dates and realistic resource commitments for your organization. Ask the participants to correct the record rather than treating silence as enthusiastic agreement. Keep unresolved issues visible with an owner and a next discussion date.

This is also a good point to check change readiness. A manager’s willingness to proceed does not establish that people have time, access or capability to participate. If the agreement relies on staff availability that nobody can protect, adjust the scope before the work begins.

Re-contract when the assignment changes

An initial agreement cannot anticipate everything. Re-contracting means explicitly revisiting the purpose, responsibilities or boundaries when the emerging work no longer fits them. It is not simply adding more tasks to the same promise.

Suppose the wholesaler discovers that the dashboard receives updates from a second warehouse with different working hours. The initial one-site investigation cannot answer the whole question. The team should describe the new dependency, its implications and the choices available: narrow the conclusion, extend the investigation with resources, or pause until the right owner is involved.

This relates to, but is not identical with, formal change control. APM’s change-control guidance describes evaluating changes to an approved project baseline and having them approved, rejected or deferred. A conversation about expectations does not replace that authorization where it is required. Follow the organization’s governance as well as updating the working relationship.

Revisit the agreement when the sponsor changes, access is withdrawn, a new group becomes affected or the requested outcome shifts. Also do it when an apparent learning exercise becomes a demand to endorse a decision already made. The change lead needs to know which role they are being asked to perform.

Record what changed and tell the people who depend on the old agreement. Updating a document in a private folder is not enough if the team still believes the original scope and dates apply. Use a proportionate communication plan for the revised commitment.

When to narrow the assignment or decline it

Sometimes the best agreement is a smaller one. If the sponsor cannot release staff time for a broad diagnosis, an initial process walkthrough may be possible. If nobody can authorize the next stage, a decision briefing may be appropriate. Be explicit about the limited conclusions either activity can support.

Other conditions call for stopping: an expectation that you misrepresent consultation, hide material findings or take responsibility for decisions you cannot control. Name the conflict plainly and ask whether the conditions can change. A signed brief does not make an unworkable relationship workable.

The approach has limits. Clear agreements cannot remove power differences, resource shortages or competing interests. This playbook combines professional guidance with an illustrative working method; it is not evidence that contracting guarantees project success. Formal employment, procurement, confidentiality or regulatory questions need the relevant qualified support.

Before leaving the first meeting, be able to complete a simple sentence: “We are investigating this problem, with these people and resources, so this person can make this decision.” If the participants give incompatible answers, the useful work is still clarification—not scheduling the rollout.