How to talk about your AI strategy to the people who actually need to hear it

We have read more AI strategy decks than is healthy. They almost all have the same problem: one deck, one narrative, intended to serve two audiences with fundamentally different concerns. The board sits through the slides looking for risk, ROI, and time-to-value. The employees who eventually read the same slides scan for whether their job changes. Neither audience gets the message they need, and both leave the meeting with the version of the story they invented in their head.

On every AI strategy engagement we run, we ship two communication artifacts, not one. A board narrative and an internal narrative. They share the same underlying strategy, but they are written for different audiences, with different framing, different examples, and different success criteria. The discipline of producing two is what protects the AI program from both board skepticism and internal panic.

What the board actually needs to hear

Board members evaluating an AI strategy are doing three things in sequence: assessing the risk of inaction, assessing the risk of action, and assessing the expected return. They are not, in our experience, interested in the technical stack. They are interested in whether the company is going to be a winner or a loser in the next two years of competitive AI deployment.

The board narrative checklist

• What is the cost of doing nothing — and who is taking share if we wait?
• What is the cost of getting it wrong — and what is our recovery path if we do?
• What is the expected return, expressed in margin or growth, with timeline?
• What are the two or three governance risks, and how are we mitigating them?
• What is the named executive who owns this, and what authority do they have?

The board version of the strategy is almost never a tour of the data stack. It is a one-page narrative that names the bet, the bet's size, the people responsible, and the risk envelope. The technical detail belongs in an appendix, not in the room.

What employees actually need to hear

Employees reading the same strategy are answering a fundamentally different question: what changes for me? They are not skeptical or eager in the abstract — they are running the calculation that everyone runs when their employer announces a new technology initiative. Will my workload change? Will my skills become obsolete? Will my team be smaller in a year? Will I be expected to learn something new on top of my current job? Will I be replaced?

The internal narrative answers those questions explicitly, in order, in plain language. The companies that handle this well do it deliberately — they don't outsource the conversation to a memo or a town hall slide. They have it directly with the affected teams, before the rumor mill has a head start. The companies that handle it badly let employees infer the answer from silence, which is the single most reliable way to generate the worst version of the story.

The internal narrative checklist

• What changes about your day-to-day work?
• What stays the same?
• What new skills are we asking you to learn, and what support are we providing?
• What is the timeline, and what are the milestones you should expect?
• What happens if the rollout is faster or slower than planned?
• Where can you ask questions without putting your job at risk?

The failure modes when one is mistaken for the other

We see two predictable failure modes. The first is using the board narrative internally: heavy on ROI, heavy on the competitive threat, light on what the employee should do tomorrow morning. The result is internal panic — employees hear the urgency and the talk of efficiency gains and assume the worst. Productivity drops, not rises, in the first 90 days of an AI rollout that uses this framing.

The second is using the internal narrative with the board: focused on training, change management, and employee comfort, light on competitive positioning and financial return. The board hears a costly internal program with no clear payoff and loses confidence in the strategy. The next budget cycle quietly trims the AI investment, and the company falls behind for reasons that have nothing to do with the technology.

A template we use on every engagement

Both narratives share a common structure: a statement of intent, the bet being made, the timeline, the success criteria, and the accountability. The framing of each element differs by audience, but the elements are the same. That symmetry is what keeps the strategy coherent even though the two audiences hear different versions of it.

Anti-patterns we cut on every AI strategy review

• A 60-slide deck that serves nobody — too long for the board, too vague for employees
• Generic 'AI will transform' language with no specific bet attached
• An internal rollout announcement that doesn't name the affected teams
• A board update that defers the financial case to 'phase two'
• Town-hall Q&A as the only feedback channel — too public to ask the real questions
• A communications plan that ends at launch — the most-needed updates come in months 2 through 6

Cadence after launch

Both narratives need recurring updates. The board version is quarterly, focused on milestones hit or missed, capital deployed, and any change to the original bet. The internal version is monthly in the first six months, focused on what is actually happening to the affected teams. Silence after the initial announcement is the single biggest predictor of program failure — it allows both audiences to invent their own story, and neither version is usually flattering.

The one-line takeaway

One AI strategy needs two narratives, written for two audiences, with two different success criteria. Get them right and the program earns both board confidence and internal trust. Get them wrong — or merge them into one — and you spend the next year managing the consequences in both rooms.