.webp)
I've been running an experiment at dinners with CIOs and business leaders across London, Frankfurt, and Chicago this past year. First, I ask about their AI strategy. The answers are impressive multi-year roadmaps, enterprise-wide deployments, thoughtful governance frameworks. Real substance.
Then I ask: "Who owns the metric for business model reimagining in your organization?"
The mood shifts almost instantly. People pause and start showing visible signs of discomfort. They start talking about innovation teams, digital transformation offices, strategy groups. But when I press on who actually has a KPI tied to their compensation for reimagining the business? Nobody.
That disconnect between the strategy we articulate and the outcomes we actually measure and reward is the entire problem with AI transformation today.
The Perform vs. Transform Checklist
Here's the pattern I see playing out across organizations. The difference between performing and transforming isn't subtle. It shows up in every aspect of how you think about AI:
How to Use This Checklist
Look at your current AI initiatives. Be brutally honest about which column they actually fall into. Not which column your strategy deck claims they're in and which column your KPIs and incentives are actually driving.
If everything is in the Perform column, you don't have a transformation problem. You have a measurement and incentive problem.
What Outcomes Really Mean
This is where language becomes a trap. Everyone uses the word "outcomes" but means completely different things.
Performance outcomes are measured against your current baseline. You reduced handle time by 40%,outcome achieved. You decreased cost per transaction by 60%,outcome delivered. These are real improvements. They make quarterly business reviews look good. But they're automation at scale. You've made an existing process faster or cheaper. The business model remains unchanged.
Transformation outcomes are measured against possibility. What customer problems can you solve now that you couldn't solve last quarter? What revenue streams exist now that were impossible before? What strategic positions can you occupy that were unavailable last year?
One measures improvement. The other measures invention.
Most organizations have carefully constructed quarterly targets and efficiency metrics. These drive perform initiatives. They're necessary. They're table stakes. Your competitors are achieving similar gains. Real competitive advantage comes from transformation from reimagining what business you're actually in.
The Measurement Gap
Here's the uncomfortable pattern I've observed across hundreds of conversations: organizations have sophisticated AI strategies but simplistic AI measurements. The strategy decks talk about transformation, disruption, reimagining. Then you look at the actual KPIs the metrics someone's compensation is tied to and it's all performance optimization.
You manage what you measure. You get what you reward.
If your KPIs are all performance metrics like efficiency gains, cost reductions, productivity improvements, that's exactly what you'll get. If someone's bonus depends on reducing operational costs by 15%, they're not going to propose eliminating the operation entirely, even if that's the right strategic answer.
But who has a KPI for:
- New capabilities deployed that were impossible last quarter?
- New customer problems solved that you couldn't address before?
- New revenue streams that didn't exist in your business model?
- Strategic options created that your competitors can't imagine?
These are transformation metrics. They’re harder to quantify because they require measuring progress toward an unclear future, not the optimization of a known present. We know how to measure efficiency; we’ve been doing it for decades. What we don’t yet know how to measure is reimagining itself how to quantify what became possible that wasn’t possible before.
So we default to what's measurable, even when we know those metrics miss the point.
The Transformation Mindset
The companies actually reimagining with AI share a pattern. They've created space for different questions:
Not "how do we optimize this process?" but "should this process exist at all?"
Not "how do we reduce cost in this function?" but "what customer value could we create if cost weren't a constraint?"
Not "how do we make our people more productive?" but "what becomes possible when AI handles the routine so humans can focus on the impossible?"
They've also changed what they measure and reward.
One financial services CEO I work with restructured their executive comp plan: 70% still tied to operational performance, but 30% now tied to launching capabilities that didn't exist at year-start.
First year was uncomfortable with lots of debate about what "counts." Second year, they had launched three new revenue streams that transformed their relationship with customers from transactional to advisory. Third year, competitors started copying what they'd built.
They're measuring progress toward a future that doesn't exist yet, rather than optimization of a present they understand completely.
How you actually make that shift—how you navigate the organizational resistance, how you run perform and transform initiatives simultaneously, how you create transformation KPIs that boards will accept—that's the deeper conversation. But it starts with acknowledging the gap between what we say we're doing and what we're actually measuring.
The Dinner Question, Revisited
So when I ask, "Who has a KPI for reimagining your business model?" I'm really asking:
Are you performing with AI or transforming with it?
Are you measuring what's easy or what matters? Are you optimizing for quarterly predictability or positioning for strategic possibility?
The silence after that question tells you everything you need to know about where most organizations actually are on the transformation journey, regardless of what their strategy decks claim.
The Great Reimagining isn't about having an AI strategy. It's about having the measurement systems, incentives, and courage to actually pursue one.
