Your team already knows your AI strategy
AUDIO OVERVIEW — Listen to this post (≈ 90 sec)
Audio overview generated with Google NotebookLM. Article written by Rick Cheever.
Most mid-market owners think they’re still deciding what AI means for their business. Their employees think the decision is already being made. They just haven’t been told what it is.
Both can be true at the same time. And the gap between those two perceptions is now the single biggest retention risk in mid-market private companies. Not AI itself. The silence around it.
The owners who come out of this period with their best people still on the team will not be the ones who picked the “right” AI strategy. They will be the ones who picked a strategy and said it out loud.
The two coherent AI strategies
There are two ways a mid-market private business can use AI strategy right now. Both can work. Both have been used to build great companies. The choice between them is not the problem. The problem is pretending the choice does not have to be made.
The substitution AI strategy uses AI to do work that people used to do. The team gets smaller. The output per remaining employee goes up. Costs come down. This is what most AI vendors are quietly selling, and it is a defensible play for businesses with thin margins or commoditized service work.
The multiplier strategy uses AI to absorb the drudgery sitting underneath skilled work, so each person can do more of what they were actually hired for. The team stays the same size, or grows. Each person’s value goes up. Wages should follow. This is the harder play because it requires real role redesign, not slide-deck redesign.
Most owners say “multiplier” in town halls and execute “substitution” in quiet vendor calls. Their employees notice. https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/the-organization-blog/are-your-people-ready-for-ai-at-scaleMcKinsey’s research suggests nearly a third of companies struggle to close AI talent and capability gaps, in part because the people closest to the work are the last to be brought into the AI strategy conversation.
The expensive mistake
The expensive mistake is not picking substitution. It is not picking multiplier either. The expensive mistake is drifting between them without naming which one is actually running.
When an owner has not decided, the team reads the drift. The marketing manager whose role gets quietly absorbed into a junior coordinator plus a Claude subscription does not need a memo to figure out what happened. The operations director who watches the CEO sit through three vendor demos in two weeks without ever being brought into the room knows that something is being discussed that they are not part of.
Top performers leave first. Not because they are afraid of AI strategy. Because they have options, and they are paying attention to which conversations they are in and which conversations they are not.
Three signals employees are already reading
Employees in mid-market companies do not need to ask the owner what is happening with AI. They are already reading three signals.
First, what gets outsourced to AI versus what gets invested in. When an entire content function moves to ChatGPT but the sales team gets a new Salesforce license and an outside coach, the message is clear about who the business is betting on.
Second, who is in the AI conversations and who is not. When a CEO is meeting with vendors and consultants and the directors who actually run the work are finding out about new tools through email rollouts, the org chart has effectively been redrawn without telling anyone.
Third, whether any role has actually been redesigned, or whether redesign is still slide talk. If after twelve months of “AI will help our people do higher-value work” no individual job description has actually changed, no one has been promoted into the freed-up time, and no one is being paid more for the new scope, the team has its answer.
What to do this quarter
Three actions an owner can take in the next ninety days to close the gap.
Pick a strategy out loud. Even a partial commitment is more useful than a vague one. “We are using AI to reduce headcount in marketing and customer service so we can invest more in sales hiring and account management” is honest. “We are using AI to free our people up” without specifying what it frees them up TO DO is not honest, even when it feels like it is.
Redesign at least one role to make it real. Pick the function where AI has actually changed the work. Write a new job description. Promote someone into it. Pay them like the new role demands. One real role redesign is worth six town halls about how AI will create opportunities.
Bring the directors into the AI conversation before the contract is signed, not after. Top performers do not need to be in every vendor call. They need to know they are in the room when the strategy is being formed. https://vistage.com/press-center/press-release/ceo-confidence-trends-downward-in-q1-2026-as-geopolitical-uncertainty-returns/The Vistage Q1 2026 CEO Confidence Index shows mid-market leaders already under pressure from tariff and geopolitical uncertainty. Adding internal trust risk to that mix is a luxury most cannot afford.
The harder honesty
The owner who says “We do not know yet how AI is going to change this business, and here is how we are figuring it out together” is in a stronger retention position than the owner who says “AI is going to free our people up to do higher-value work” without any plan to back it up.
Honest ambiguity beats confident vagueness. Especially with the people the business cannot afford to lose.
