Rethinking Work Series: Part 2
Imagine a company automates most of its contracting process. Work that previously took 20 days now takes three. Some positions disappear, others change, and employees begin managing AI agents that perform activities once handled by entire teams.
Six months later, the company conducts its usual performance reviews. But what exactly is it evaluating? The number of contracts an employee processed? Their productivity? Their individual goals? And how do those measures connect to the organization's actual business results?
We're investing billions in changing how work gets done, yet much of our approach to measuring performance still assumes work is performed by individuals operating within relatively stable roles and organizational structures.
I believe we need to rethink performance around three connected dimensions: business performance, organizational effectiveness, and individual performance. I think of them as a three-legged stool. Each matters, but none gives us the complete picture on its own.
Business Performance: Are We Creating Value?
Let's return to our contracting example. We've reduced turnaround time from 20 days to three. That's an impressive productivity improvement, but has it translated into better business results? Are we closing more deals? Increasing revenue? Reducing costs? Improving customer experience? Are we managing commercial risk effectively? If contracts are moving faster but sales conversion hasn't improved, we've made a process more efficient without necessarily improving business performance.
This is where I'd start measuring success: not with how much work AI performs, but with the business outcomes that work is supposed to produce.
Organizational Effectiveness: The Missing Connection
This is where I think the most significant opportunity lies. Organizations have traditionally measured financial results, operational efficiency, employee engagement, and individual performance through different systems. But AI is making the relationships between them more important and more complex.
In our contracting example, I'd want to know how well the entire system operates. How much work is automated? Where are decisions getting stuck? Are approval rights clear? Are people using the technology effectively? Are AI agents producing reliable results? Perhaps we've automated contract review, but managers still insist on approving every agreement. Or sales teams are bypassing the system because they don't trust it. We've improved the technology without fixing the organization.
Organizational effectiveness connects business outcomes to the way work actually gets done, including the interaction between people, processes, decisions, technology, and AI agents.
We can increase productivity without improving organizational effectiveness. And we can improve organizational effectiveness without necessarily achieving the business outcomes we intended.
Understanding those differences tells us where to intervene.
Individual Performance: What Are We Actually Evaluating?
This is where traditional performance management becomes increasingly challenging.
Consider a commercial lawyer who previously reviewed 30 contracts a month but now oversees AI agents processing 500. Is that person suddenly a dramatically higher performer?
What if those contracts contain errors? What if the lawyer identifies a systemic risk that prevents a significant financial loss? What if the contracts are processed perfectly, but the company still isn't closing more business?
Measuring the volume of work no longer tells us enough about the individual's contribution. As AI takes over routine activities, employees will increasingly be responsible for judgment, problem-solving, managing exceptions, improving workflows, and making decisions. We need to evaluate how well they perform those responsibilities, not simply count the activities completed by the technology they oversee.
But we also need to be careful about accountability. An employee shouldn't be judged solely on business outcomes influenced by pricing, market conditions, leadership decisions, or technology beyond their control.
The challenge is understanding individual contribution within a system where results are increasingly produced by people and AI working together.
The Three Legs Must Work Together
Now imagine looking at these three dimensions together. Business performance tells us whether we're achieving our goals. Organizational effectiveness helps explain how our operating model is producing those results. Individual performance tells us how people are contributing and where capabilities need to improve.
If revenue isn't growing despite faster contracting, we can investigate whether the issue is sales effectiveness, pricing, customer demand, or the contracting process itself.
If employees are producing significantly more work through AI but quality is deteriorating, we can examine the technology, the workflow, and the judgment of the people overseeing it.
And if business results are improving while employee performance ratings remain unchanged, perhaps we're measuring the wrong contributions.
The value isn't in creating three new scorecards. It's in understanding the relationships between them.
What If We Started With the Business, Not the Employee?
This raises a more fundamental question about the future of performance management.
What if we stopped treating individual employee performance as the starting point and instead began with the outcomes the business needs to achieve? We could work backward to understand the processes, decisions, technology, AI agents, and human capabilities responsible for producing those outcomes.
That would require connecting information that currently lives in different systems: financial results, CRM data, operational workflows, customer experience, AI activity, and employee performance.
It wouldn't eliminate the need for individual performance management. We'd still need clear expectations, feedback, development, and accountability. But those practices would become part of a broader understanding of how the organization creates value.
AI could also help us identify patterns across these systems, revealing where performance is constrained by processes, decisions, technology, or capabilities rather than assuming the problem lies with individual employees.
Of course, data quality, privacy, and the limits of attributing outcomes to specific causes would matter enormously. The goal isn't to monitor every employee or turn every activity into a metric. It's to give leaders a more complete understanding of performance so they can make better decisions.
The Future of Performance Management Is Bigger Than Performance Management
We've spent decades refining individual goals, ratings, feedback, and performance reviews. Those practices still have a role, but they were never designed to explain how an entire organization performs. And as AI changes the relationship between work, people, and business outcomes, that limitation becomes harder to ignore.
I believe the opportunity is to move beyond systems that primarily measure employee performance toward systems that help us understand organizational performance, including the contributions of people, technology, and AI. That doesn't mean abandoning individual accountability. It means putting it in context.
In my previous article, I argued that AI gives us an opportunity to fundamentally redesign how organizations operate. But we can't transform how work gets done while continuing to measure success through systems designed for the organization we left behind.
