74% think they're transforming. 20% actually are.
This is the worst part.
Not the gap between where your company is and where it should be.
The gap between what you think you're doing and what you're actually doing.
Deloitte surveyed 2,700 executives across industries about their AI transformation. They found something uncomfortable:
34% of organizations report being in "deep transformation" with AI. Sounds ambitious.
Then they looked at what "deep transformation" actually meant in those organizations:
- They have an AI strategy document
- They've allocated a budget
- They've hired some AI people
- They've started pilots
- Everyone talks about it
Sounds like transformation. But here's the reality check:
Only 20% of that 34% actually show meaningful business impact.
The Gap Between Talk and Results
What's the difference between the 34% and the 20%?
Process redesign.
The 34% add AI on top of existing processes. They had an approval workflow — now an AI pre-screens it. They had a customer data sheet — now AI fills 70% of it. Same workflow. More automation.
The 20% actually ask: "Why do we have this process?"
Then they demolish it and rebuild it around what AI enables.
So instead of "AI approves, human reviews," they ask "do we need approval at all?"
Instead of "AI fills the form, human verifies," they ask "do we need this form?"
That's where the real gains come from. Not the technology. The reorganization.
The 14% in the Middle
These are the organizations that are doing something. But they're also doing it wrong.
They're in the dangerous zone: they've invested enough money to be committed, but not enough to actually transform. So they're stuck in a hybrid state.
Worse, they're creating resistance. Employees see: "The company bought this AI thing. They're asking us to learn it. But they haven't changed anything about how we work. So I have to do the old job plus learn the new thing."
That's exhausting. And it's the best way to kill adoption.
Why 80% Fail
Because most organizations try to do AI transformation without doing organizational transformation.
They add a Chief AI Officer (new hire, expensive). They set aside a budget (money, no strategy). They run pilots (safe, isolated, not connected to the real work).
Then they expect results.
What actually happens: The pilots work fine in their bubble. Then they try to scale them, and they break. Because the rest of the organization isn't built to work that way.
It's like getting a Ferrari and expecting it to go faster on a dirt road.
What the 20% Do Differently
1. They start with outcomes, not tools: "What problem are we solving?" not "How do we use AI?"
2. They redesign first, automate second: Before you implement AI, you've already changed the process.
3. They measure real impact: Not "we have X AI projects running," but "revenue per employee increased by Y% and here's why."
4. They accept disruption: They know some jobs will change fundamentally. They plan for that, don't hide from it.
5. They move fast: The pilots aren't perfect. But they launch them into the real organization and learn from reality, not PowerPoint.
The Real Question
Are you in transformation, or are you in theater?
Theater looks like: lots of announcements, committees, strategic initiatives, consultant reports, town halls about AI.
Transformation looks like: processes that stopped existing, jobs that got redesigned, revenue that moved up, and no big announcement about it.
The organizations winning keep quiet and keep shipping. The ones failing keep talking and keep tweaking.
This is part of my series on the future of work. In each post I take a specific research finding, prediction, or move — and ask what it means about the organizations we live in.
If you're managing through transformation, here's the real work: The Little Book for New Managers.
Want to read the next one?
More organizations getting it wrong (and a few getting it right).
That clears it up, Lior
Want more on managing in the AI era?
The Little Book for New Managers has helped 30,000+ managers figure out what actually works.
Get the book ($19) →