Episode Insights ·
Enterprise AI Strategy: The 3-Legged Stool Every IT Leader Needs
The Technology Vanguard with Bob Hansen

Every IT leader is being asked the same question right now: what is our AI plan? Lyza Valenzo, now VP of Global IT at Oliver Healthcare Packaging, answered it on The Technology Vanguard with one of the clearest enterprise AI strategy frameworks we have heard. She calls it a 3-legged stool, and the part most companies skip is the seat.
Lyza did not start in technology. She spent years in customer service, procurement and transportation before a CIO pulled her into IT. Later she grew the IT team at eCore from about 6 people to 23, moved the company to Office 365 within days of COVID, rolled out Salesforce and led IT through several acquisitions. That business-first view shapes how she thinks about AI.
Why most enterprise AI strategies start as trial and error
Lyza is honest about where most organizations are. IT leaders are looking across the table at their peers and asking, "What are you doing?" Most strategies today are some mix of experimentation and trial and error.
That is not a weakness, as long as there is a structure around the experiments. Her framework gives you that structure, and it works for any company, in any industry.
The 3-legged stool of enterprise AI strategy
Think of your AI adoption framework as a stool. Three legs hold it up. The seat is what makes it something you can actually sit on.

Leg 1: Put AI in everyone's hands
Get an enterprise AI tool out to the whole organization, whether that is Copilot, ChatGPT or another approved platform. People are already using AI as consumers. They need the same chance at work to build prompt skills and find time savings in their day.
Leg 2: Cultivate a group of power users
Find the small percentage of people who are savvy, curious or in roles where AI will change everything. Give them close access to IT, elevated licenses and room to experiment with agents and new tools. Her advice to them: "Try it. Don't wait for permission. Go try it."
Lyza believes this group becomes your next generation of leaders, both leaders of people and leaders of technology. They are the proof that IT can drive the business, not just keep the lights on.
Leg 3: Pick slam dunk use cases
Look for painful, problematic processes in the business. Then start with the ones that are easy wins. Early successes build experience, credibility and momentum before you take on the harder problems. Vendor partners like Microsoft and CDW may also offer people, funding or credits to help you get traction.
The seat: digital competency in the fabric of the organization
This is the part of the enterprise AI strategy conversation Lyza rarely hears anyone talk about. The seat is digital competency across the whole company.
That means working with HR business partners and the executive team to bring digital curiosity into hiring and performance reviews. How willing is this person to use technology? Are they tech averse? When curiosity becomes a shared language, AI and digital transformation stop being an IT project and become part of how the organization works.
Why "move fast and fix it later" sinks AI projects
When Bob asked for the worst advice Lyza ever received, her answer was clear: move fast now and fix it later. "You never fix it later," she said.

The mess lands on the team. Instead of working on new technology and growing their careers, people spend their time unraveling shortcuts. That is how you lose motivation and lose good people.
AI makes this lesson even more urgent. Lyza points to master data. If your data is not clean, the quality of your AI suffers, and so do your KPIs. As she put it, your KPIs become "junk in, junk out," and you cannot make sound business decisions on them. Master data governance is tedious work, but it pays off.
When failing fast is the right call
Planning does not mean analysis paralysis. A later boss taught Lyza to "fail fast": spot a bad path quickly, make the decision and cut your losses. Her rule of thumb is simple. Small batches can move fast. Big, foundational initiatives get planned.
How to balance speed with a solid foundation
Every IT leader knows the moment: the business wants it done in 30 days. Lyza's approach to that pressure comes down to transparency and partnership.

- Give options. Bring several paths, not a single yes or no.
- Show the risk and reward. Cover cost and the impact on people for each option.
- Split the project. When a project feels like boiling the ocean, break it apart. Move the low-risk pieces quickly in an agile way and give the risky pieces the time they need.
- Decide together. Socialize the options with the executive team so the decision is top down and shared. Nobody is left on an island, and nobody gets blamed when a hard project gets hard.
Governance and cybersecurity still apply
Lyza recommends a steering committee for AI, along with a business continuity plan, a risk register, identity management and clear rules for confidential versus public information. Your technology has to enforce those policies, not just your behavior policies. On cybersecurity, ask every AI vendor how security is built into their tools. Her bottom line: "You can't not do it because it's going to create risk. How are you going to mitigate the risk?"
AI governance: NIST AI Risk Management Framework.
The leadership lesson behind the strategy
Lyza's framework works because she leads with people first. She credits a 10-month leadership program, the Society for Information Management's Rising Leaders Forum, with changing her life, and she has since sent three of her own team members through it.
Her advice for anyone early in their career fits the same pattern:
- Volunteer for the thing others avoid. The big, gnarly projects build your confidence and your reputation.
- Take lateral moves. "Taking lateral positions or business positions is not a step back." Her years in the business are why she can talk to procurement, logistics and production planning teams in their own language.
- Notice mentorship moments. "You don't need a mentor to be mentored."
Watch the full episode
Hear Lyza Valenzo and Bob Hansen go deeper on AI strategy, data, cybersecurity and leadership on The Technology Vanguard.
Frequently asked questions
What is the 3-legged stool of enterprise AI strategy?
It is Lyza Valenzo's AI adoption framework. The three legs are putting an enterprise AI tool in everyone's hands, cultivating a group of power users, and starting with slam dunk use cases. The seat is digital competency built into hiring, reviews and company culture.
Why is master data important for AI?
AI is only as good as the data behind it. Poor master data lowers the quality of AI outputs and makes KPIs unreliable, so leaders end up making decisions on "junk in, junk out."
Is "move fast and fix it later" ever okay?
For small batches, yes. For big, foundational initiatives, Lyza recommends planning up front, because the fix rarely happens and the team is left cleaning up the mess.
How should IT leaders respond when the business wants to move faster?
Offer options, show the risk and reward of each, split large projects into fast and careful pieces, and make the final call together with the executive team.
Do you need a technical background to lead IT?
Not necessarily. Lyza came from customer service, procurement and transportation. Her business experience and people leadership are what brought her into IT leadership.
Keep exploring
What should we ask our next guest?
Got a question for a future guest on The Technology Vanguard? Here is how to Ask the Vanguard:
- Open the latest episode on YouTube.
- Leave a comment that starts with "Ask the Vanguard:"
- Watch for your question in an upcoming interview.
New episodes drop every week.

THE VANGUARD INSIDER
Steal the Playbooks of Top Tech Leaders
Every week Bob Hansen sits down with CIOs and IT leaders who've done it for real. Join the Vanguard Insider and get their sharpest lessons in your inbox, no full episode required.
- The 5-minute recap of every new episode
- Leadership lessons you can use Monday morning
- Early access to new episodes and guest announcements
- Behind-the-scenes questions that didn't make the cut
