AI Will Change More Than Work, It Will Change How We Measure Good
Yesterday I had the opportunity to speak with a couple of AI experts about where the technology is headed. The conversation was fascinating, but it also reminded me of something I experienced firsthand years ago — a lesson about automation that feels incredibly relevant today.
AI is changing work and the metrics we use to evaluate it. We need to adjust those metrics to understand what’s actually happening inside our organizations.
A Lesson from Early Automation
About 15 years ago, I worked on an automation initiative at an insurance carrier.
We automated some of the simplest transactions in the contact center, things like:
- Address changes
- Replacing a vehicle on a policy
- Updating lienholders
These were quick, repeatable, and low-complexity tasks. The logic at the time was straightforward: if we automate these tasks, productivity should increase. Instead, something unexpected happened. Productivity dropped dramatically and employee dissatisfaction rose. The numbers looked terrible and leaders were confused.
Through research, we discovered that when we automated the easiest transactions; the remaining work became dramatically harder.
The employees who used to handle a mix of simple and complex interactions were suddenly left with only the most difficult ones. Calls took longer, problems were harder to solve and the emotional intensity of the role increased.
And with all that, we were still measuring performance using the same productivity metrics that assumed a balanced mix of work. We had changed the work, but not the measurement.
AI Will Accelerate this Dynamic
Fast forward to today. AI is poised to remove even more routine work across insurance, from underwriting to claims to customer service. Low-level tasks will increasingly be handled by automation and AI systems. Humans will step in primarily for the hardest problems. Which means something important will happen:
- Average handle times will rise
- Throughput may drop
- Cases will look more complex
And if leaders are still using the same productivity metrics, they may make the wrong conclusions. They may believe performance is declining when, in reality, the nature of the work has fundamentally changed.
The New Leadership Question
This is why the real leadership challenge of AI may not be adoption, but measurement. The old metrics were built for a world where humans handled everything. If AI removes the simplest work, then we have to ask new questions:
- What does productivity mean when humans only handle complex cases?
- How do we measure the value of judgment, empathy, and expertise?
- What does success look like for an adjuster or underwriter working alongside AI?
Redefining “Good” in the Age of AI
In the AI era, organizations may need to shift toward metrics that better capture complex human work, such as:
- Resolution quality, not just speed
- Customer outcomes, not just throughput
- Expert decision-making, not just task completion
- Employee cognitive load and sustainability, not just volume
Because when humans are left with only the hardest problems, their value changes. And so should the way we measure it.
The Leaders Who Get this Right
AI will transform workflows across industries, and the organizations that benefit will rethink how they define success. Twenty years ago, I watched productivity “fail” because our metrics hadn’t caught up to automation. As AI reshapes work again, leaders have an opportunity to learn from that mistake. The real question isn’t just how AI will change the work, it’s how we change to measure good.
