Change Management: Communications — The Parallel Track That Starts Before the First Workflow Is Redesigned

This article is part of a 27-article series on the AI Business Transformation Methodology. This piece introduces the change management function, makes the case for proper resourcing, and covers the communications workstream in depth as the first and most urgent leg of change management.

Your Level 1 triad has set the strategy. The Level 2 portfolio has been built. Domain owners have been chartered. The capability decomposition is about to begin.

And somewhere on the fourth floor of your finance building, a senior analyst who has been with the company for twelve years just heard from a colleague that “AI is going to redesign how accounts payable works.” Nobody from leadership has told her anything. She does not know what this means for her job. She does not know who decided this, when it is happening, or whether she should be updating her resume. What she does know is that she has read about the wave of AI-attributed layoffs across the industry, the companies that cut thousands of positions citing AI capabilities that had not been deployed yet, and her employer just chartered something that sounds a lot like what those companies did before the cuts started.

This is what happens when the transformation work begins without the communications strategy. The work itself may be excellently conceived and well-intentioned. The people affected by it are filling the information vacuum with fear.

Before addressing what to do about the information vacuum, it is important to understand where communications sits within a broader discipline that the methodology treats as a parallel track running alongside the entire transformation.

Traditionally, change management has three legs. Each activates at a different point in the transformation lifecycle, and each serves a distinct purpose.

The first leg is communications: building awareness of why the transformation is happening, managing the narrative, and keeping every level of the organization informed as the work progresses. This is the leg that activates first, before the capability decomposition, before the education cascade, before a single workflow is redesigned. It is the subject of this article.

The second leg is org impact assessment and job redesign: analyzing what the redesigned workflows mean for roles, skills, organizational structure, and career paths. This leg activates as workflow designs produce output, because you cannot assess how jobs change until you can see the redesigned workflows. Article 15 covers this in depth.

The third leg is training: building the system-specific competence that people need to work effectively in the new workflows with the new technology. This leg activates at Level 4, when the technology is selected, configured, and ready for people to learn. You cannot develop system-specific training materials, job aids, or step-by-step guides until the systems exist. Article 22 covers this.

These are not sequential phases that happen one after another. They are parallel workstreams that activate at the point in the lifecycle where they become relevant, and the first two run concurrently for much of the Level 3 work. The domain owner who treats change management as something to address after the design work is finished is the domain owner whose transformation will be technically sound and organizationally rejected.

Every experienced change management professional has managed communications through an ERP implementation, a platform migration, or a digital transformation. The methodology described in this article draws on the same research-grounded discipline that has guided those programs for decades. But AI transformation introduces communications challenges that are qualitatively different from anything the profession has encountered, and a communications strategy that does not account for those differences will underperform regardless of how well executed it is.

The second difference is the external narrative. When your organization launched its last digital transformation, your employees were not going home at night and reading headlines about the new platform eliminating their profession. AI transformation operates inside a media environment saturated with stories about job displacement, workforce reduction, and human obsolescence. Your communications strategy is not operating in an information vacuum. It is competing with an external narrative that is often more alarming than the internal reality. As Article 1 documented, many of those headlines describe organizations that cut positions citing AI capabilities that had not actually been deployed. But the fear they generate is real, and it is present in your workforce before the first internal communication is sent.

The fourth difference is the absence of a stable end state. Typical transformation programs have a go-live date after which the system stabilizes and roles settle into a new normal. AI capabilities evolve continuously. The capability meta-layer tagging described in Article 12 exists precisely because new capabilities will emerge and reshape what is possible at each workflow step. When that happens, the roles defined against those steps change too, as Article 14’s reengagement trigger mechanism addresses. The communications strategy cannot promise employees a stable future state that will hold. It must instead build the organizational capacity for continuous adaptation, which is a fundamentally different communications objective than preparing people for a one-time transition.

These six differences do not invalidate the proven change management methodology that follows in this article. They add a layer of complexity and sensitivity that the methodology must account for. Every deliverable, every message, every sender decision should be evaluated through the lens of these AI-specific dynamics. The change management professional who applies standard OCM methodology without adjusting for these factors will produce a technically sound communications plan that fails to address what people are actually experiencing.

The research on this point is unambiguous and has been consistent for years.

There has not been a more important time to get this right. The scope of AI-driven business transformation, redesigning workflows, redefining roles, rebuilding how entire functions operate, represents a level of change that most organizations have never attempted at this scale. And the six AI-specific dynamics described above make the communications challenge harder, not easier. This is the one function where we strongly recommend organizations err on the side of overbudgeting rather than risk the consequences of underinvesting in the very capability that determines whether the rest of the transformation is adopted.

This is the nuance that most organizations miss, and it is the nuance that separates the organizations that build momentum from those that build resistance.

Communications does not begin when the domain owner starts executing their charter at Level 3. It begins at Level 2, when the Level 1 triad announces the transformation portfolio and charters domain owners. That announcement is the first communication event for the enterprise.

This is the moment when the broader organization learns that transformation is happening, which domains are in the first wave, and what the overall strategic direction is. The message at this stage is enterprise-level: here is why we are transforming, here is the competitive and strategic context that makes this necessary, and here is what you can expect to see in the coming months.

In an AI transformation, starting early is even more critical than in previous transformations because of the external narrative problem. Your employees are already anxious about AI before you announce anything. The enterprise-level communication at Level 2 is the organization’s first opportunity to distinguish its specific transformation from the fear-driven narrative employees are consuming externally. The message must be concrete enough to be credible and honest enough to build trust: here is what we are doing, here is why, and here is how we intend to treat the people who will be affected by it. The workforce investment premise from Article 1 is not a philosophical position. It is a communications strategy.

This Level 2 communication is not the detailed, domain-specific communications strategy. It is the enterprise-level awareness that creates the foundation on which domain-level communications builds. Article 4 describes the chartering process where domain owners receive their imperatives. That chartering moment should include the enterprise-level communication that prepares the broader organization for what is coming.

When the domain owner receives their charter (Article 9), one of their first actions should be activating the domain-level communications strategy. This is more specific than the enterprise announcement: it addresses the specific function being transformed, the specific people affected, and the specific timeline.

The domain owner does not develop this strategy personally. Change management practitioners, chartered by the domain owner as part of the transformation team, develop the strategy and the detailed plans that implement it. But the domain owner is accountable for ensuring it is chartered, resourced, and integrated into the transformation timeline from the beginning, not bolted on after the design work is underway.

Who delivers the domain-level messages follows the same preferred sender principle from the research. The domain owner and VPs deliver the strategic messages: why this domain is being transformed, what the business outcome will be, and how this connects to the enterprise strategy. Directors and managers deliver the personal impact messages: what this means for our team specifically, what to expect in the coming weeks, and what support is available. The change management practitioners prepare both groups to deliver those messages effectively, drafting the key messages, coaching the senders on delivery, and scheduling the communications to align with transformation milestones.11

In an AI transformation, the coaching of preferred senders takes on additional importance. The domain owner and VPs need to be prepared not just to deliver the strategic rationale but to address the AI-specific anxieties their teams are experiencing. Directors and managers need to be equipped to have honest conversations about what AI means for specific roles on their teams, including the ability to say “I do not know yet, but here is when we will have more clarity” rather than offering false assurances. The change management practitioners should prepare sender-specific talking points that anticipate the AI-specific questions employees will ask, because the questions in an AI business transformation are different from the questions in other transformations.

Understanding that communications is important is the easy part. Knowing what the communications workstream actually produces, at the deliverable level, is what separates organizations that manage change from organizations that talk about managing change.

The deliverables fall into two categories: foundational deliverables that are produced once and updated throughout the transformation, and ongoing deliverables that are produced and refined continuously as the transformation progresses.

The stakeholder analysis is the first deliverable the change management practitioners produce, because everything else depends on it. It identifies every group affected by the transformation, the degree to which each group is impacted, their specific concerns, and how they prefer to receive information. In an AI transformation, the stakeholder analysis should also map each group’s current relationship with AI: are they already experimenting independently, are they anxious and disengaged, or are they somewhere in between? Research has documented distinct adoption personas within organizations, from enthusiastic maximalists to cautious observers to active resisters, and the communications strategy needs to address each differently.

The change impact assessment analyzes how each specific change affects each stakeholder group. As the transformation progresses through capability decomposition (Article 10) and workflow redesign (Article 12), the impact assessment is updated with increasingly specific information about what is actually changing. In an AI transformation, the impact assessment must address not just process changes but the shift in the nature of work itself: from execution to orchestration, from data entry to judgment, from sequential tasks to exception management. These are identity-level changes, not just process-level changes, and the communications strategy must acknowledge the difference.

The communications strategy is the overarching plan that governs all subsequent communications activities. It defines the audiences, the messaging themes for each audience, the channels through which communications will be delivered, the frequency of communications, the preferred senders for each message type, and the metrics by which communications effectiveness will be measured. This is not a newsletter schedule. It is the strategic architecture for how information flows to every affected person throughout the transformation.

Key messages by audience are developed and updated throughout the transformation lifecycle. Different audiences need different messages at different times, delivered by different people. The change management practitioners draft messages tailored to each stakeholder group, aligned to where that group is in the change process, and prepare the appropriate senders to deliver them. This is an ongoing deliverable because the messages evolve as the transformation produces new information about what is changing and what it means for people.

The communications calendar maps specific messages, delivery dates, audiences, channels, and senders across the transformation timeline. It is aligned to transformation milestones from the broader Level 3 work: capability decomposition workshops, education cascade sessions, workflow redesign reviews, and cross-department coordination checkpoints. Every significant transformation milestone should have a corresponding communications activity that prepares people for what is happening and what comes next.

In an AI transformation, the two-way channels serve an additional function: surfacing the invisible adoption that is already happening. McKinsey’s finding that employees are experimenting three times more than leaders realize4 means the listening channels need to actively seek information about how people are already using AI tools, what they are learning, and where they are encountering risks. This ground-level intelligence is valuable not just for communications but for the entire transformation: it reveals where adoption is happening naturally, where governance gaps exist, and where the organization’s assumptions about readiness are wrong.

The substance of what is communicated changes as the transformation progresses. This progression is not arbitrary. It maps to Prosci’s ADKAR framework, which identifies five sequential elements that individuals need to successfully adopt a change: Awareness, Desire, Knowledge, Ability, and Reinforcement. Communications primarily serves the Awareness and Desire elements, with later phases supporting Knowledge and Reinforcement.14

In an AI transformation, the Awareness content must do something it has never had to do in previous transformations: directly counter an external narrative that is shaping employee perceptions before the organization says a word. The awareness messages must acknowledge the broader industry conversation about AI and workforce displacement, connect it honestly to what the organization is and is not doing, and establish the workforce investment premise from Article 1 as the organization’s governing principle. Employees who have been reading about AI layoffs for months need to hear their own organization’s leaders say, clearly and specifically, “that is not what is happening here, and here is why.”

As the transformation progresses into capability decomposition and the education cascade, communications shifts to building Desire: why should I engage with this, what is in it for me, and how will this make my contribution more valuable rather than less? In an AI transformation, the Desire-building messages must address the identity dimension. It is not enough to explain that the new workflow is more efficient. The communications need to convey that the evolved role is more strategic, more analytical, and more interesting than the current one. The human contribution in the redesigned workflow is not a diminished version of what the person does today. It is an elevated version, one that focuses on judgment, exception handling, relationship management, and the kind of contextual decision-making that AI cannot perform. This is the message that converts anxiety into engagement.

As workflow redesign and job redesign produce outputs, communications supports Knowledge: what specifically is changing, what do the redesigned workflows look like, what are the three categories of role evolution (augmented, consolidated, emergent from Article 15), and what does this mean for specific teams and individuals? This is the most sensitive phase of the communications lifecycle and the one where the preferred sender framework matters most. Employees want to hear about how changes affect them personally from their direct supervisors, not from a company-wide email.

At Level 4, when technology is deployed, communications supports Ability and Reinforcement: how to do the new work, where to find help, what early successes look like, and what the organization is learning from the first iteration cycles. The training workstream (Article 22) carries the primary weight at this phase, but communications continues to reinforce the narrative and celebrate progress. Because AI capabilities evolve continuously, the communications function at Level 4 and beyond must also prepare people for the reality that the current state is not the final state, building the organizational capacity for continuous adaptation that the no-stable-end-state dynamic requires.

One challenge that is unique to AI transformation is ensuring that the communications narrative is technically accurate. The redesigned workflows from Article 12 include AI-enabled steps that most employees and many managers do not fully understand. The communications strategy does not need to explain the operational details of how AI works at each workflow step. That is the responsibility of the training workstream at Level 4 (Article 22) and the org impact assessment practitioners during job redesign (Article 15). But the communications strategy does need to ensure that the key messages, the sponsor talking points, and the awareness-level framing are accurate enough to be credible and honest enough to build trust.

The CAIO’s AI-Business Translators, described in Article 5, serve as subject matter experts for the change management practitioners developing key messages and sponsor preparation materials. The Translators understand both the technical capabilities and the business context well enough to ensure that the narrative the organization tells about its AI transformation is grounded in reality. They are not the communications senders. They are the partners who help the communications team ensure that when the CEO says “AI will transform how our finance organization operates,” the statement is accurate, and when the domain owner stands in front of a town hall and fields questions about what AI means for the department, the talking points reflect what AI can and cannot actually do.

This partnership is particularly important for three communications responsibilities. First, building awareness-level messages that are specific enough to be credible without crossing into training-level operational detail. The communications should convey that the finance organization’s workflows are being redesigned to leverage AI for higher-speed, higher-accuracy operations, with human roles evolving toward judgment, exception handling, and strategic analysis. That framing is accurate, builds awareness, and does not attempt to deliver the operational training that belongs in Article 22. Second, preparing sponsors to field AI-specific questions honestly. When employees ask “Will AI replace my role?” the domain owner needs a prepared, accurate answer that the Translators have helped validate. Third, correcting misinformation. When the two-way feedback channels surface misconceptions about AI capabilities, whether overblown fears or unrealistic expectations, the Translators help the communications team craft corrections that are technically sound and delivered through the preferred sender framework.

Communications activates across multiple levels of the methodology, with each level’s communications building on the previous one.

At Level 2, enterprise-level awareness communications accompanies the portfolio announcement and domain owner chartering. The CEO, CSO, and CAIO deliver the strategic rationale to the organization, including the direct acknowledgment of the AI-specific dynamics that make this transformation different from previous ones. This prepares the ground for everything that follows.

At Level 3, domain-level communications activates when the domain owner receives their charter and runs continuously alongside all Phase 3A through 3C work described in article 9, Domain Owners Charter. The content evolves as the transformation progresses through capability decomposition, education cascade, workflow redesign, data readiness, cross-department coordination, and job redesign. The change management practitioners produce the deliverables described in this article and keep them current as the transformation produces new information.

At Level 4, communications evolves to support technology deployment: preparing people for go-live, explaining what the new systems and tools do, and reinforcing the narrative during the transition period when first-cycle friction is expected and budgeted for, as Article 12 described.

At Level 5, communications becomes part of the continuous transformation operating rhythm. As the organization builds the capability for ongoing self-optimization, the communications function shifts from transformation-specific messaging to an embedded practice of keeping the organization informed about continuous improvements, new capabilities, and evolving roles.

The thread that runs through all of these levels is the workforce premise from Article 1: this transformation is investing in people, not eliminating them. Every communication, at every level, should be consistent with that premise. The moment the communications say one thing and the organization’s actions say another, trust collapses, and trust is the foundation on which everything else in the methodology depends.