Why Learning Measurement Has Failed, and How to Fix It

The Learning Industry Is Reaching a Turning Point

Walking through the halls of ATD 2026 in Los Angeles recently, one thing became impossible to ignore. The energy around learning measurement, ROI, business impact, and proving value felt fundamentally different this year. The sessions were packed. The conversations repeatedly circled back to the same themes:

• “How do we prove learning is making a difference?”
• “How do we connect learning to business outcomes?”
• “How do we build confidence with stakeholders?”

Even in the ATD bookstore, the books flying off the shelves were overwhelmingly focused on ROI, analytics, learning measurement, and business alignment. For me, that doesn’t feel like a coincidence. I think it reflects something much bigger happening across our industry.

Our Current Flawed Paradigm

For years, many learning teams have relied heavily on activity metrics to demonstrate value:

This doesn’t mean that I believe learning leaders are asleep at the wheel when it comes to learning measurement. I think deep down, most learning leaders already know they don’t fully capture the value learning creates inside an organization. I think the issue is not intent, but rather lacking the tools and training to truly understand and measure what truly matters for business stakeholders when it comes to the business impact of learning.

I also believe that this lack of tools and training has led to a paradigm that measuring and proving actual business impact – especially learning spend Return on Investment (ROI) – is impossible. For example, I recently had a senior learning leader from a large organization say to me
“You can’t measure learning ROI, so you shouldn’t even try.”

Reading between the lines, it boils down to this viewpoint: others have tried and failed, so it’s not something that’s realistic to attempt.
I fundamentally disagree.

The “You Can’t Measure ROI” Myth

Human progress has always been built on tackling problems that were initially considered too difficult, too complex, or too ambitious. The first attempts at flight failed. The first attempts at space travel failed. The first attempts at mapping the human genome failed.

Yet nobody concluded that these pursuits were impossible and therefore not worth attempting. Instead, each failure improved understanding, refined methodology, and moved us closer to a workable solution.

Measuring learning’s contribution to business performance is no different. The fact that previous approaches have struggled does not mean the challenge is impossible. It simply means we have more work to do.

The Missing Middle

The more conversations I had at ATD, the more I found myself returning to the same thought. Learning leaders are not struggling because they don’t care about business impact.

Quite the opposite. They’re struggling because they can clearly see where investment goes in and they’re increasingly being asked to explain the value that comes out. What they can’t clearly see is everything in between. Over time, I found myself returning to the same idea. There is a part of the learning journey we simply haven’t been able to see.

We’ve Been Asking The Wrong Question

The reason I disagree is because I think we’ve been trying to solve the right problem in the wrong way. For years, the conversation around learning measurement has revolved around one big question: “What is the ROI of learning?”

On the surface, that sounds entirely reasonable. After all, every other business function is expected to demonstrate value. Why should learning be any different?

The problem is not the question itself. The problem is that we’ve often tried to jump straight to the answer. We attempt to draw a direct line between a training course and a financial outcome. Revenue, cost savings, productivity gains, customer outcomes and return on investment all become the focus of the conversation.

Those are questions that CFOs care about. They’re questions executives care about. And they’re absolutely questions that learning leaders should care about too.

But here’s the challenge: Learning rarely creates business outcomes directly. People do.

A sales training program doesn’t increase revenue. It’s salespeople applying what they learned that increases revenue.\

A leadership program doesn’t improve employee engagement. Instead, it’s leaders changing how they lead that improves employee engagement.

A customer service course doesn’t improve customer satisfaction. It’s employees applying new skills and behaviors that improves customer satisfaction.

That’s an important distinction. Because once you see it, you start to realize that learning doesn’t create value at the point of consumption. It creates value at the point of application. And that’s where I believe we’ve been looking in the wrong place. We’ve spent years measuring what happened during learning.

We measured:

• Who attended.
• Who completed.
• Who participated.
• Who was satisfied.

But business value is rarely, if ever created during learning. Business value is created afterwards. It’s created when somebody does something differently because of what they learned. When their capability improves. When their confidence grows. When better decisions are made. When their execution becomes more consistent. The business value of learning is, at its heart, really a very simple calculation:

When behaviors change leads directly to improved performance.

That’s where the value lives. And I believe that is the tension our industry is now beginning to confront. Because executives are increasingly asking a different question.

Not: “How much learning did we deliver?”

But: “Did it make a difference?”

That’s a fundamentally different conversation. And it’s one that requires a fundamentally different way of thinking about learning measurement. Because once you accept that learning doesn’t create business outcomes directly, another question naturally emerges: “How does learning actually create value?”

That’s a surprisingly important question. For years, many of us have treated learning and business outcomes as though they sit side by side. But for most of us, the reality is that learning happens and is tracked, and the business is influenced somewhere down the line.

And somehow we’re expected to prove a connection between the two.

But the reality is far more nuanced. Business outcomes don’t suddenly improve because somebody completed a course. Something happens in between. In fact, a lot happens in between. When somebody learns something new, they may develop new knowledge, skills, or capability. As capability grows, confidence often grows alongside it. As confidence grows, people become more willing to apply what they’ve learned. As application increases, behaviors begin to change. Those behavioral changes influence how work gets executed. Improved execution influences performance. And improved performance is what ultimately drives business outcomes.

The more I thought about this, the more I realized that perhaps the challenge facing our industry isn’t measuring business impact. Perhaps the real challenge is understanding everything that happens before business impact. The problem isn’t that learning value is invisible. The problem is that we’ve been looking for it in the wrong place. Because if we can’t see whether capability improved, whether behaviors changed, or whether performance increased, then proving business value becomes incredibly difficult. Not because the value isn’t there. But because the evidence trail is missing.

Stepping Inside the Missing Middle.

The more I explored this idea, the more I realized that value doesn’t appear all at once. It accumulates through a sequence of human performance improvements that occur between learning and results.

When I visualize it, it looks something like this:

What becomes interesting is that most learning systems are very good at measuring the beginning of this chain. And most organizations are reasonably good at measuring the end of it.

What we’ve historically struggled to understand is everything happening in the middle. Yet that’s precisely where learning either succeeds or fails. It’s where capability is built. It’s where confidence develops. It’s where behaviors change. It’s where performance improves. And it’s where the evidence trail connecting learning to business outcomes is either created or lost.

Looking at learning through this lens changed my thinking entirely. It helped me realize that learning measurement isn’t really about proving ROI at the end of the journey. It’s about understanding and validating the chain of events that creates ROI in the first place. And that’s where I believe Return on Expectation (ROE) enters the conversation.

Why Return on Expectation Matters

This is where I believe the conversation around Return on Expectation (ROE) becomes increasingly important. Not as a replacement for ROI. And certainly not as an excuse to avoid measuring business impact. Quite the opposite. I believe ROE exists because learning leaders need a more credible way of understanding and validating the chain of events that ultimately leads to business outcomes.

Because when you think about it, most organizations already measure the beginning of the journey. They know what learning was delivered. They know who attended. They know who completed. And most organizations are reasonably good at measuring the end of the journey as well. They can see revenue. They can see productivity. They can see customer outcomes. They can see operational performance.

What has historically been much harder to understand is everything happening in between.

• Did capability improve?
• Did confidence improve?
• Did people actually apply what they learned?
• Did behaviors change?
• Did performance improve as a result?

I see ROE as the missing evidence trail between learning activity and business outcomes. It helps us move beyond simply reporting what learning happened and towards understanding whether learning actually transferred into workforce improvement. Because if learning doesn’t create value at the point of consumption, but rather at the point of application, then surely our measurement approaches need to evolve as well.

We need to become better at understanding whether learning is changing capability. Whether capability is influencing behavior. Whether behavior is improving performance. And whether those performance improvements are contributing to organizational outcomes. That doesn’t mean every learning initiative can be perfectly measured. It doesn’t mean every outcome can be attributed with scientific precision. But it does mean we can build far greater confidence than we’ve historically been able to.

And I believe confidence is ultimately what learning leaders have been missing. Not confidence that learning matters. Most organizations already believe that. Confidence in understanding how learning creates value. Confidence in demonstrating that value to stakeholders. And confidence in making better decisions about where future learning investments should be made.

That’s why I believe ROE matters. Not because it replaces ROI. But because it helps explain the journey that creates ROI in the first place.

Making the Missing Middle Visible (A New Era for Learning)

For years, learning leaders have been asked to explain business value without being able to clearly see the journey that creates it. We’ve measured learning activity. We’ve measured business outcomes. But we’ve often lacked visibility into everything happening between the two.

The Missing Middle has always existed. We’ve simply never been able to see it clearly. I believe that is beginning to change. Not because we’ve suddenly discovered a perfect measurement framework. But because we’re starting to better understand where learning creates value, how that value moves through organizations, and what evidence it leaves behind.

If we can make The Missing Middle visible, we may finally be able to build a more credible connection between learning investment and business value. And that may be one of the most important opportunities facing our industry today.

The future of learning measurement isn’t about proving that learning happened. It’s about understanding whether learning made a difference. And perhaps more importantly, understanding how it made a difference. Because that’s where the evidence trail lives. And that’s the conversation I want to explore next.

This is the conversation Scott will be continuing live at TICE 2026. Catch his session, “From Learning Data to Business Proof: What Most L&D Teams Are Still Missing,” on June 18 at 9:45 AM ET in Raleigh, NC. Stop by booth 29 to keep the conversation going.

Or, connect with our team directly, now!