Boards are asking for AI roadmaps. Vendors are pitching new copilots. Leaders are announcing “AI-first” futures at town halls.
Meanwhile, most managers have never been trained on how to use AI in their own workflow, let alone how to coach their teams on it. The result is predictable. You get noise, experiments, pockets of brilliance, and very little sustained, measurable change.
Gallup just released new research that makes this painfully clear. It should be a wake-up call for every executive who thinks “rolling out tools” equals “driving adoption.”
At PerformanceLabs, this is the gap we live in every day. We run AI workshops with sales, marketing, and customer teams across media and SaaS.
The pattern is the same: AI strategy at the top, AI experimentation at the edge, and a giant missing middle where managers should be.
It is time to stop pretending that an AI strategy matters if your managers and frontline teams are flying blind.
Nobody owns the day to day behavior change. Managers are not equipped with real examples, prompts, and coaching moves. Reps are not given simple, measurable workflows where AI is the default, not the exception.
The hard truth: the tools are not the problem
Gallup highlights a stat that should stop any AI steering committee in its tracks. A 2025 MIT NANDA study found that only about 5% of organizations report measurable ROI from their generative AI investments.
Five percent.
This is not a tooling problem. The tech is already more powerful than how most companies are using it.
Gallup’s own data backs this up. Even as AI availability has surged in the past two years, access has not translated into adoption or ROI. The bottlenecks sit squarely in human and managerial space: unclear, local use cases and resistance from middle managers and frontline employees.
In other words, companies bought the race car and never taught anyone to drive it.
What employees are really telling you about AI
Here is where it gets even more frustrating.
When Gallup asked employees about the biggest barrier to AI adoption, the top answer was not “we do not have tools.” It was “unclear use case or value proposition,” cited by 16% of respondents. Legal and privacy concerns came next at 15%, followed by “lack of training or necessary knowledge” at 11%.
For employees who do not use AI at all, the story is even sharper:
44% say they do not use AI because they do not believe it can help with the work they do.
Only 16% say the main issue is lack of access to AI tools.
Others point to resistance to changing how they do their job (11%), lack of knowledge on how to use AI (10%), or feeling unsafe using AI (8%).
Read that again. Nearly half of non users are not blocked by IT or budget. They simply do not see how AI connects to their day to day work.
That is a leadership failure, not a software limitation.
Nobody has translated “AI strategy” into “this is how AI helps you write better outreach, build smarter proposals, run cleaner one on ones, and protect your time.
The missing piece: managers who actually lead AI use
Gallup’s core finding is simple and powerful.
Within organizations that are investing in AI:
Employees who strongly agree that their manager actively supports their team’s AI use are 2.1 times more likely to use AI a few times a week or more.
They are 6.5 times more likely to strongly agree that the AI tools provided are useful for their work.
They are 8.8 times more likely to strongly agree that AI gives them more opportunities to do what they do best every day.
That is not a small lift. That is the difference between “we have licenses” and “this actually changed our performance culture.”
Yet only 28% of employees in organizations that have started implementing AI strongly agree that their manager actively supports their team’s use of it.
So three out of four employees are basically on their own.
Executives keep asking why adoption is slow. The answer is sitting right in front of them. Managers are the compound interest of AI adoption, and most are being left out of the plan.
Gallup’s four best practices are necessary, but not sufficient
Gallup outlines four practices that correlate with higher AI usage:
Communicate a clear strategy for AI integration.
Champion AI use at the team level.
Provide role specific training that maximizes value and mitigates risk.
Establish clear policies and guidelines for responsible use.
All four are spot on.
Where most companies fall down is in execution. They write the strategy. They publish the policy. They might even do a one time training.
Then they stop.
Nobody owns the day to day behavior change. Managers are not equipped with real examples, prompts, and coaching moves. Reps are not given simple, measurable workflows where AI is the default, not the exception.
How we connect the dots at PerformanceLabs
Our view is blunt: AI adoption is a behavior change problem disguised as a technology initiative.
The organizations that win do three things differently:
1. They start with real workflows, not abstract use cases
In our AI workshops, we do not start with “what is generative AI.” We start with “show me your calendar and pipeline.”
How are you prospecting?
How do you prepare for calls?
What kills your time between meetings?
Then we layer AI directly into those workflows with live reps, not just slides. When a seller sees a messy account plan turned into clean next steps in 30 seconds, adoption stops being theoretical.
2. They turn managers into AI coaches, not traffic cops
Most AI programs treat managers like a compliance layer. “Make sure your team follows the policy.”
That misses the point.
We work with managers first. They learn hands on how to use AI to plan their week, run better one on ones, and coach using actual call notes and deal data. Once they feel the value personally, it becomes natural to model AI use and advocate for it with their teams.
We then give managers concrete tools:
Team specific prompt playbooks.
Scorecards that track AI assisted workflows.
Simple coaching questions that unlock usage, like “where did AI save you 30 minutes this week.”
This turns Gallup’s insight into real behavior. Managers become the multiplier Gallup is describing, not a missing variable.
3. They measure behavior, not just licenses
You cannot manage what you refuse to measure.
Instead of tracking “number of users provisioned,” we help leaders measure:
Percentage of key workflows that are AI assisted.
Time saved on specific tasks like prospecting, email drafting, and pre-call research.
Pipeline lift tied to AI supported activities.
Why frustration is justified, and what to do next
If you are an executive who has invested in AI and you are not seeing results, your frustration is warranted.
But the answer is not another tool, another task force, or another strategy memo.
The answer is to own the human side of this.
Admit that your AI strategy is incomplete without a manager strategy. If your plan does not explicitly say how managers will use, model, and coach AI, it is not a plan.
Fund role specific training, not generic “AI 101.” Teams need to see AI in the context of their actual work. They need prompts, examples, and safeguards built around real workflows.
Make manager behavior visible. Start asking every leader how they are using AI to run their team. If it is not on their scorecard, it will not show up in their calendar.
Treat AI adoption as ongoing, not a launch event. Habits take repetition. Make AI a standing part of team meetings, one on ones, and performance reviews.
At PerformanceLabs, we help companies do exactly this. Our AI workshops, prompt playbooks, and coaching programs are built to move you from “we bought tools” to “our managers and teams use AI confidently, safely, and in ways that show up in the numbers.”
Gallup’s research is clear. Manager support is the single biggest lever for AI adoption and perceived value.
The question is not whether you have an AI strategy.
The question is whether your managers and frontline teams have any idea how to live it.