A change communication plan people can actually use

A useful change communication plan starts with the questions people need answered to do their work. It connects those questions to an accountable person, an appropriate moment, an accessible channel and a way to respond. A calendar of announcements is part of the plan, but it is not the plan itself.

Imagine a repair service replacing scattered emails with a shared request queue. Coordinators will assign work differently, technicians will record progress in a new place and managers will review different information. This is a fictional example throughout. Sending everyone a launch email might announce the change accurately while leaving its most important practical questions unanswered.

Start with the work that is changing

Describe the proposed difference without the campaign language. “From Monday, eligible daytime requests will enter one shared queue; the duty coordinator will assign them” is something people can question and rehearse. “One team, one future” may be a slogan, but it does not tell a technician where an urgent request belongs.

Separate what has been decided from what is still being designed. State what remains unchanged, what people are being asked to do and when further decisions will be made. If the exception route is unresolved, say who will resolve it and when the answer is due. Do not fill an operational gap with optimistic wording.

The OASIS planning framework begins with objectives and audience insight before moving to strategy, implementation and evaluation. Applied to this workplace example, that means defining the communication task before choosing a channel. “Technicians can explain where an urgent request goes” is a more useful objective than “send three newsletters.”

Use your readiness assessment to identify issues that messages cannot fix. Missing system permissions require access to be arranged. A workload conflict requires a decision about capacity. An explanation may help people understand those problems; it cannot resolve them on its own.

Organize audiences around their questions

A stakeholder map identifies who may be affected. The communication plan needs a more practical next layer: what does each group need to know, when, and in what working circumstances? Start from stakeholder conversations, not assumptions based on job titles.

In our repair service, coordinators need to know who owns an unassigned request. Technicians need to know where to record a temporary repair and how to get help without delaying urgent work. Managers need to know what the queue’s status fields do—and do not—say about performance. A colleague returning from leave needs a way to catch up without searching through a month of announcements.

Check access as well as information needs. Is someone away from a desk for most of the day? Are there different shifts, languages or assistive-technology requirements? Which existing channels do people actually use? Do not turn audience segmentation into personality labels such as “resistant old-timers.” The useful distinction is a difference in work, impact or access that changes the plan.

The government inclusive communications template asks planners to consider audience needs, channels, alternative formats and testing early. For our example, a short shift briefing with a readable reference guide may reach people whom an intranet article misses. Test that choice with the people concerned rather than assuming that a printed sheet or a video is accessible to everyone.

Make the core message candid and usable

Draft a short common explanation covering the reason for the change, the intended difference, the consequences for the audience and the available support. Keep the underlying facts consistent while changing the level of practical detail for each role. Consistency does not require identical messages.

For the fictional repair service, a draft could read: “Requests currently arrive in several inboxes, making responsibility difficult to track. We will trial a shared queue for ordinary daytime requests in one location. The duty coordinator will assign those requests. Urgent work keeps its current escalation route during the trial. We will review the experience before deciding whether to expand.”

That message makes a limited commitment. It does not promise an easier day for everybody or imply that the trial has already succeeded. It also creates questions worth answering: which requests are ordinary, who covers the coordinator’s absence, and where will the review appear?

Maintain a small decision log alongside the frequently asked questions. Date the answers and identify their owner. An unanswered question should have a status and next update, not disappear because it makes the launch page look untidy. If an earlier answer changes, show the correction where people saw the original.

Give each communication a job and an owner

The compact plan below is an illustrative starting point, not a completed project schedule. Set real dates only after the operational decisions and preparation are credible.

Moment and audience	Question to answer	Format and accountable owner	Feedback route
Before the trial: coordinators	Who assigns work and covers absences?	Walkthrough led by the service owner	Record unresolved handoffs for a decision
Before first use: technicians	Where do ordinary and urgent requests go?	Shift briefing and reference guide, owned by the local manager	Private or group questions to a named contact
During the trial: all affected roles	What changed after the first problems?	Dated update from the trial lead	Check whether the revised instruction is usable
Before expansion: affected teams	What did the trial show, and what remains uncertain?	Review discussion led by the decision owner	Publish the decision and explain the reasoning

A channel is not an accountable person. “Intranet” cannot verify an answer about access, and “the change team” may mean several people each assuming someone else will respond. Name the role responsible for the accuracy of the message and the role handling questions. They may be different people.

Prepare managers before asking them to brief their teams. Give them the current explanation, an opportunity to challenge it and a route for questions they cannot answer. A cascade becomes unreliable if each manager has to invent missing policy. Equally, avoid requiring every local question to wait for a central newsletter when the local owner already has authority to resolve it.

Do not confuse listening with participation

Tell people which decisions they can influence. Asking for comments on wording is different from inviting them to design the workflow. Neither is automatically wrong; the problem is presenting a settled decision as an open consultation.

CIPD’s employee-voice guidance treats voice as both expression and influence, and emphasizes listening and action rather than a single feedback initiative. For this plan, make the route from question to response explicit. Who reads it? Who can decide? How will the contributor learn what happened?

Consider a technician who asks how to record a temporary repair that still needs a replacement part. The communication owner should not answer with “remember to keep the queue up to date.” The question exposes a missing operational distinction. The service owner needs to agree the status and responsibility; the guide then needs to change; affected colleagues need the corrected explanation.

Keep a short record: issue, affected work, owner, next update and resolution. Where a suggestion cannot be accepted, explain the reason and any alternative. Offer a route for concerns people cannot comfortably raise in a group. Do not promise anonymity unless the mechanism genuinely supports it, and do not publish personal details in a general question log.

If the same concern returns repeatedly, ask whether the rule or objective needs reconsidering. Our comparison of single- and double-loop learning explains the difference between clarifying an instruction and questioning the assumption behind it. Repeating a message more loudly is not always the appropriate response.

Check understanding, not just distribution

Record whether a message reached the intended audience, but keep that separate from whether it was understood or usable. An email open is not agreement, and agreement is not the ability to carry out a task. The distinctions in ADKAR can help organize follow-up questions without turning the model into a score.

The government Evaluation Cycle distinguishes communication outputs from audience responses, outcomes and wider impact. In our repair service, those levels might be the briefing delivered, the route understood, eligible requests handled as intended, and a service experience worth sustaining. Evidence at one level does not automatically establish the next.

Ask someone to talk through a realistic exception or find the relevant instruction. Observe where the explanation breaks down, then separate a wording problem from a process problem. If the technician can explain the route but cannot access the queue, another briefing is unlikely to help. If a manager interprets “closed” differently from a coordinator, settle the definition rather than counting page views.

Use feedback to adjust timing and channels. Preserve a record of what changed so that later evaluation can distinguish the original plan from the version people actually experienced. Avoid attributing any service improvement solely to communication when training, staffing and the workflow changed at the same time.

When this plan is not enough

This is a practical synthesis of institutional communication guidance, not a tested intervention with a guaranteed effect. The example plan is invented and needs local adaptation. The government frameworks were written for government communication; the role-based repair-service application is an editorial adaptation, not an official template.

High-consequence or sensitive changes may require specialist communication, employee-representation processes and other formal arrangements beyond this guide. A campaign plan must not be treated as a replacement for them. Nor should the plan be used to turn reasonable disagreement into a messaging defect.

A good final check is simple: can an affected person find the current answer, understand what it means for their work, raise a concern and discover what happened to it? If one of those paths is missing, add that path before adding another announcement.