substack.com - The Senior Manager’s Guide to AI From Hype to Real Impact - Part 2
How to Drive Team AI Adoption
Section titled “How to Drive Team AI Adoption”Welcome back to my series on driving AI adoption as a senior leader. In Part 1, I focused on personal use because you cannot “talk the talk” if you don’t “walk the walk”. Those were the first two stages of the SPARK framework: Start and Partner. If you’re still finding your footing with AI, read Part 1 first.
Today I will talk about Level 3: Activate your team. In the next and final installment, I’ll cover the org-wide shifts: replicating best practices across functions and kindling a true AI-first transformation. Together they form SPARK: Start, Partner, Activate, Replicate, Kindle. You will have an AI campfire burning in your company in no time. 😉
Subscribe if you want to get the third part in your mailbox.
Level 3: Activate Your Team
Section titled “Level 3: Activate Your Team”How fluent is your team in AI today? Are they debating the necessity of MCP servers, do they keep a chatbot open like it’s their second brain, or are they still at square one?
Most teams I’ve seen look like a bell curve: a few pioneers far ahead, a few lagging behind, and the bulk somewhere in the middle. Your role is to spot the trailblazers and use their knowledge and energy to pull the rest forward.
So how do you actually do this?
AI may feel new, but adopting it follows the same patterns as any change initiative. So here we can get help and ideas from the established models (ADKAR, Bridges’ Transition, McKinsey 7S and many more). My favourite is Kotter’s 8-step framework because it’s practical and easy to apply.
Kotter’s eight steps:
- Create a sense of urgency
- Build a guiding coalition
- Form a strategic vision
- Enlist a volunteer army
- Enable action by removing barriers
- Generate short-term wins
- Sustain acceleration
- Institute change
Step one is urgency. Your team needs a reason to care. It might be the thrill of working on the cutting edge. It might be fear of missing out. Or it might be a top-down mandate. Shopify’s CEO, for example, made headlines when he declared that ” reflexive AI usage is now a baseline expectation ”. Business schools have drilled the ” burning platform ” mantra into generations of managers. With AI, the platform is on fire: adopt it or risk being left behind. Which is exactly how I feel when my kids start talking GenAlpha language at the dinner table.
Next comes the guiding coalition. At the company level, this usually means executives with budget and authority. At the team level, that’s you. (Yes, you’re the coalition. Congratulations.)
Third is vision. People need to see what they are working towards. For instance:
- Productivity vision: “We’ll free up 10 hours a week by cutting admin and reporting, so we can do more interesting work”.
- Customer-centric vision: “We’ll serve customers faster and better, with fewer errors.”
- Removing barriers vision: “We can finally do things ourselves instead of waiting on others.”
My own team is already seeing the first signs of that last one.
A project turned “red” after the partner team building our solution was pulled onto a different priority. Dead end, or so it seemed. At an Agentic AI training, one of my product managers discovered an internal agent with a key capability we needed. She built a quick prototype, proved it could work, and found a tech team willing to scale it. Just like that, we were not “red” anymore.
Fourth, enlist your volunteers. Aim for 5–10% of your team to act as AI scouts. Their job: explore, experiment, and return with maps for the rest of the group. They are the cornerstones of adoption, and they need both recognition and support.
Five, enable action and remove barriers. This is where tactics matter.
Start with training. Point your team to high-quality, time-efficient options. For example, Andrew Ng’s AI for Everyone (free, six hours course on Coursera) is a solid foundation. LinkedIn Learning, Google, and Vanderbilt also offer good courses, and your company will likely build internal trainings as well. The key is curation: give people a clear path so they don’t waste time wandering. For instance: foundation → advanced prompting → agentic AI.
Next, give your team explicit permission to carve out time to learn and experiment. Share a vetted list of tools they can safely try and build with.
I put a recurring Friday morning block on everyone’s calendar called Experiment with AI*. At first, people (myself included) were skeptical. We are professionals — we can block our own time, right? But it worked. Now when someone wants to go deeper on AI, they say, “Let me take this up on Friday.”*
Expectation from the boss helps. A shared anchor in the calendar helps. For a distributed team, like mine, having a shared timeblock helps.
Even 90 minutes a week, consistently, compounds into real knowledge.
Weave AI into your team’s existing rhythms. Add “AI wins & learnings” to your weekly huddle. At first, the agenda slot may sit empty. Don’t worry, within a few weeks you might find the entire meeting consumed by AI stories. What gets attention, gets adoption.
Also create spaces for problem-solving: a Slack channel, office hours, or a shared doc where people can post blockers. Giving the team a place to get unstuck keeps momentum going.
Kotter’s sixth step is about short-term wins. One of your AI scouts will soon find a “killer use case” — something that makes their work faster or better. Spotlight it. Ask them to run a Show & Tell, record a video, or write a how-to guide. Share it widely and celebrate.
Start looking for use cases where AI can help your team:
- High inbound questions? Build a persona trained on your knowledge base.
- Endless reporting? Automate the pulls and formatting.
- Repetitive tasks? Explore agentic AI tools for automation.
Each win proves the value and builds momentum.
At this stage, quality starts to matter. You don’t want “workslop” — outputs that look polished until you scratch the surface and find nonsense underneath. For example, one of our partner teams is working to automate weekly reporting with AI. The first drafts looked promising, until we noticed it was inventing new retail subcategories out of thin air. To avoid this, set clear standards and run quality checks. Otherwise, you’ll trade speed for confusion.
Step seven is about sustaining acceleration. Don’t stop after the first wins. Once your team has skills and confidence, push into harder use cases. Bring everyone together and brainstorm: what takes the most time? What creates the most frustration? A group complaining session, handled well, can turn into a goldmine of automation ideas.
This stage, along with Kotter’s final step, institutionalizing change, leads directly into Levels 4 and 5 of the SPARK framework. So let’s pause here and recap Level 3.
The hallmarks of Level 3:
AI tools are part of daily work.
A shared prompt library exists.
An AI persona answers common user questions.
At least one process is automated through an agent.
Every week, a new AI learning or win surfaces in team discussions.
How to level up:
- Share wins beyond your team. As manager, amplification is your job.
- Ask your team to produce assets others can use, like recordings, prompt libraries, how-to guides, even simple apps.
- Dedicate some of your time to marketing these wins. Spreading the word turns local gains into organizational progress.
Interlude - How to Deal with AI Fear and Resistance
Section titled “Interlude - How to Deal with AI Fear and Resistance”At some point you will face resistance, fear, competition and disillusionment.
Resistance. When there is too much change too fast. When the expectations from leadership are too high, but no help is offered. When there is fear of identity loss (X job will get automated and disappear!). People don’t like change imposed on them. They want to feel in control and be in the driving seat. Engage them early in the change. Invite concerns and ideas. Let them solve some of the problems themselves. Help remove barriers and show the way.
Fear. People fear that machines will take our jobs. Machines fear that people will unplug them. Everyone had trust issues. Can we just set this straight? If machines didn’t take our jobs, we’d still be farming twelve hours a day, six days a week. Technology has always shifted work and improved quality of life. Perhaps AI will mean that our kids work twenty hours a week and live better than we do today.
The reality is: jobs are constantly being reinvented. Try explaining to your grandma what you do for a living.
We can shape the change, or it will be shaped for us.
Imagine shopping in a few years. Agents will handle the entire process. Want organic local produce? Your agent sources it and delivers it on Saturday morning. On a budget? The agent finds discounts and meets your target price. Humans will still be in the loop: designing software, fixing hardware, preventing disruptions, improving customer experience.
Change is like a river. Swim against it, and you’ll exhaust yourself. Swim with it, and the current carries you forward.
Appeal to self-interest. People with the latest skills will land jobs, launch companies, and create value. In the early internet days, anyone with basic HTML and CSS could make a living building websites. AI will create similar openings. If you’re learning now, you’re in the global top 10%. That’s a good place to be.
Competition: the “Hunger Games” version of AI. People may hoard their best tricks, hoping to look more valuable than colleagues when the next re-org hits. History shows this is a losing strategy. In every tech wave — electricity, computers, internet — the long-term winners weren’t the lone wolves. They were the builders: those who created systems, taught others, and expanded what their organizations could do. Better to grow the pie than fight over crumbs.
As leaders, our role is to make knowledge sharing the currency that counts. Celebrate it. Create “force multiplier” awards. Make AI-savviness a visible promotion criterion. When people see that teaching others pays, they’ll do more of it.
Finally, disillusionment. The AI vision is inspiring, but the road there is bumpy. I keep hearing from my colleagues: “This feels too much like coding”. Which is true, except coding never also wrote you a haiku about cheese.
Pilot projects will stall. This is normal. At first, change is exciting: new ideas, fresh energy, leadership attention. Then come the roadblocks, the slipping dates, the impatient leaders. Rosabeth Moss Kanter called it the law of change: “Everything looks like a failure in the middle”.
When you find yourself in the middle, do not despair. Adjust expectations, find workarounds, keep learning, and push through. Every successful transformation has passed through this stage.
How is your team AI adoption going? What tactics or rituals have worked best for you? Leave a note in the Comments below.
***
That’s it for today! In Part 3, we’ll move to the final two levels of SPARK: Replicate best practices across the organization and Kindle an AI-first transformation.
Subscribe so you don’t miss it.
Until next time and happy AI exploring! 🙂