Actual AI adoption doesn’t start with a strategy presentation or a big rollout announcement — that’s only the introduction. It begins with leaders showing AI’s possibilities to their teams in the simplest, most visible ways (like a rewritten email here or a meeting summarized with AI notes).
For Matt Berg, Director of AI Business Solutions, Business Strategy at Microsoft, this isn’t theory — it’s daily practice. In his conversation on #shifthappens, Matt shares how small, consistent actions from leaders can turn AI from a curiosity into a trusted part of the workplace. The real magic starts beyond launch and through execution.
Lead by Doing, Not Just Talking
Matt’s first advice for leaders is direct: Use AI out loud.
That might mean saying at the start of a meeting, “I had Copilot reread the notes from our last call, and something jumped out at me that I totally missed.” Or sharing that Copilot helped rewrite a message in a fraction of the time it would have taken otherwise. These everyday touchpoints send a stronger signal than any formal training because they make AI feel normal and not separate from the work.
Prompting Is Just Good Leadership
When Microsoft first introduced Copilot, Matt’s team had a clear idea of who would excel at using it, and their reasoning seemed solid at the time.
"In the early days when Microsoft launched Copilot and we had this early adopter program, we had a hypothesis that the people who would do best with Copilot, who would write the best prompts, were our younger, early-career people because they get tech [and] they're digital natives," Matt recalls. "It turns out we were dead wrong. Our best prompters are the managers, front-line managers."
It wasn’t about technical fluency at all. It was about communication. The managers who excelled weren’t necessarily power users, but they knew how to convey intent, break down tasks, and set expectations. Those same skills translated directly into better AI outputs.
The best results come from prompts that set context, define expectations, and outline what’s needed — exactly how a good manager briefs their team. Short, vague prompts often produce weak outputs, and clear, detailed ones unlock real value.
This process reframes AI adoption as less about developing technical skills and more about extending existing strengths. Here’s how to put them to work.
1. Use Prompts That Set the Tone for Your Team
Great leaders give clear direction, and AI responds best to that same clarity. Treat prompting as an extension of your leadership style — how you set expectations, frame objectives, and define outcomes. When leaders model thoughtful prompting, they give their teams permission and inspiration to do the same.
2. Turn AI Curiosity Into a Daily Habit
You don’t have to block an hour for “AI time” on the calendar. The real shift happens when AI slips naturally into the day: asking a quick question, rewriting a draft, summarizing a thread before a meeting. These micro-moments, done in the open, build trust far faster than a one-off training session.
3. Spark Momentum Through Micro Wins
Matt has seen how a single, low-stakes challenge can light up adoption. That might be having everyone in a meeting generate an AI image, rewrite a tricky email, or pull an unexpected insight from a dataset. “We had everybody generate an AI image. That's it... on everyone's phone, they went to the link, they went directly to Copilot, and they selected their own prompt, and they said, ‘generate me an image.’ Guinness verified it. So yeah, I am a Guinness Book of World Records holder because I was there,” he shares.
4. Lower the Barrier So Everyone Can Try
Not every team will have immediate access to full Copilot licenses, and that’s okay. As Matt points out, most Microsoft users can already access Copilot chat for free with a secure Microsoft account. It won’t tap into your company’s data, but it’s perfect for building familiarity, showing what’s possible, and encouraging safe experimentation.
5. Make It Real Before You Scale It Big
AI doesn’t have to transform your entire organization overnight. Sometimes the best starting point is personal: use it to speed up your own work, improve your own clarity, or explore new insights into your own data. Those visible wins become proof points others can rally around, and the culture will follow.
Measuring ROI Without Overcomplication
For leaders under pressure to justify investment, Matt keeps the math simple. A Microsoft 365 Copilot license costs $30 per user per month. If someone saves just one hour of work that month, the tool has already paid for itself.
In practice, the return often comes much faster — sometimes within hours. However, ROI involves saving time and, more importantly, creating value. A leader who saves time drafting reports can reinvest that time in higher-impact work like coaching or strategy.
AI as a Personal Productivity Game-Changer
Matt is especially enthusiastic about the Copilot analyst agent. Instead of wrestling with formulas or waiting for an analyst, he can now describe what he wants in plain language — and Copilot builds it out, complete with charts, graphs, and even Python scripts.
“I’m obsessed with it,” he admits. “I can throw together a spreadsheet, ask which three customers I should go after next, and it just… does it. That’s something we didn’t have when we launched Copilot, but it’s a game changer.”
These personal, everyday wins are powerful proof points for leaders. When they share them, they help others see AI not as a distant corporate initiative but as a tool they can use today.
From Curiosity to Confidence
The thread running through Matt’s approach is simple: AI adoption happens when leaders take the first step, use the tools openly, and create space for their teams to follow.
Whether rewriting a prompt during a meeting, gamifying an experiment, or showing a time-saving Excel insight, every action builds trust and familiarity. Over time, these moments turn AI from something new and intimidating into something routine, like checking email.
“The reality is you can’t know everything,” Berg says. “But there’s nothing wrong with that. Use AI to help you be smarter, faster, and your team will see it’s okay to do the same.”
The path to AI confidence isn’t paved with big launches — it’s built on small, visible acts of leadership. If you’re ready to move your team from curiosity to confidence, start by modeling the behavior you want to see.
Episode Resources
#shifthappens Research: AI & Information Management Report
#shifthappens Insights:
- Gen Z and Digital Fluency: Is Your Workplace Culture Ready?
- Empowering Career Growth in a Hybrid World: Strategies for Leaders to Thrive
#shifthappens Podcasts:
- How to Lead in the Age of AI
- How AI Transforms Copywriting
- Leadership Principles That Drive Big Bet Transformation
- Digital Transformation and Tech Insights
- People First, Tech Second: Leading with Kindness in Times of Digital Transformation
Matt Berg on LinkedIn
Dux Raymond Sy on LinkedIn
Mario Carvajal on LinkedIn
