Fewer AI initiatives. Every one proved in the accounts.
I teach leadership teams a method for choosing AI work that moves the numbers the board reads, and for showing that it did. Your people learn it on live proposals and keep it after I leave.
I've sat in the CEO seat. I help PE owners and portfolio leadership make the AI adoption decision deliberately, price the dependency before it's accepted, and build so a vendor problem can't take the business down with it.
AI adoption success is still the exception not the rule.
62%
EY
42%
S&P Global - 2025 Survey
80%+
RAND
The GenAI Paradox
89% of organizations use AI and 80% report individual productivity gains, yet 60% to 94% see zero measurable P&L impact.
A Framework for Identifying Value Opportunities
AI initiatives rarely fail because the model wasn’t smart enough. They fail because the money goes in before anyone agrees which result should move, where the work is held back, or how success will be measured.
Almost none of them fail because the AI model wasn't smart enough.
They fail because AI is only about 10% of the equation. The remaining 90% is data architecture, workflow integration, change management, and operational plumbing. If you focus only on the model and technology stack, you're setting yourself up to join the failed column, with an expensive exercise and no tangible returns.
What is possible when AI is done well?
WHAT YOU LEAVE WITH
AI work that shows up in the numbers, and the evidence to prove it.
Your board wants AI to move the results. Your team needs a way to choose the right work and show that it delivered.
Your leadership team is being asked where to invest in AI, what return to expect and how anyone will know. I teach them a method for answering those questions themselves, using your own priorities and the proposals already on the table.
By the end, every proposal your team has worked through will have:
A financial result the board recognizes, and a baseline set before any money is spent.
A tested answer on whether AI is the right fix, and which simpler options it beat.
A named owner, a governance plan and the skills it needs, agreed before approval.
A measure that shows whether it worked, and a point at which it gets stopped if it doesn’t.
Your team leaves with a ranked set of initiatives it can defend, and a method it can run again whenever priorities change.
HOW IT WORKS
From board priorities to a verified result
A five-phase sequence. Nothing gets chosen until the evidence says where it belongs.
01
Understand
The board sets its financial priorities. The work behind them is traced to the one step holding results back, and a baseline is set.
WHO: C-Suite, Boards
02
CHOOSE
Simpler fixes are tested first. AI goes forward only where it beats the best non-AI option.
WHO: Executives and Managers
03
Design and justify
The new workflow, its governance and its business case are built at full cost.
WHO: Organization
04
DECIDE
Initiatives are gated, ranked against capacity and approved in a single portfolio session.
WHO: Organization
Understand What AI Changes
AI Advisory for Boards
Make decisions you can defend.
Independent advice, education and challenge for boards assessing AI's impact on their business, their investments and their future value.
THE BOARD'S PERSPECTIVE
01 Where is our business exposed?
02 Which investments deserve support?
03 What evidence demonstrates value?
Unsure about where to start?
Additional Resources
A CLEARER APPROACH TO A COMPLEX REALITY
Deterministic by default.
AI by advantage.
Resilience by design.
Every use of AI buys a capability and creates a dependency on a model, a provider, and infrastructure the business doesn't control. The scarce skill right now isn't AI fluency. It's judgment: knowing where that trade is worth making, and building so the dependency can fail without the company failing with it.
This isn't an argument against AI. The businesses that win the next decade will use it aggressively. It's an argument against adopting it by default instead of by decision.
UNLOCK THE POTENTIAL OF AI WITH CONFIDENCE
My Books
