Business leaders have long been asking L&D to prove its value. To justify the cost. For a long time, the best option was a report of learning completions.
But the truth is, standard learning activity metrics (completions, time spent, satisfaction scores) don’t answer the right question. They never have. The only question they answer is whether learning content was delivered to employees and completed by them.
The question leaders have always needed to answer is more consequential: Did the workforce get better at something that matters to the business?
With agentic AI on the rise, the speed at which a workforce can adapt is everything. Your business can only move as fast as your workforce can learn. Yet, according to ATD, less than one third (30%) of organizations are good at using learning program data to make business decisions, probably because that data is hardly connected to the business in the first place.
That’s starting to change. New technical capabilities and a growing body of skill intelligence are finally making it possible to connect learning to business outcomes that leaders care about. Industry leaders gathered at Degreed Focus London 2026 to discuss these changes, how they’re playing out in real organizations today, and what’s next for personalized activity metrics.
The Problem with Measuring Learning Activity in L&D
Learning activity is easy to generate and measure. What’s difficult is connecting that activity to business-critical workforce capability outcomes. At Focus London, Lori Niles-Hofmann, Co-Founder of 8levers, said that L&D teams can string these metrics together to paint a bigger picture. But that doesn’t always mean it’s the bigger picture. The standard metrics are usually flimsy support for a meaningful conclusion.
“We could look at data-driven learning design and aggregate measurement to where we could start saying, ‘Maybe X plus Y equals Z,’ and we can put things together but it was sketchy,” said Niles-Hofmann.
She proposed something better: closed-loop reporting, or the ability to “micro-isolate different types of learning interventions” to identify their impact in real time. For example, if an employee was preparing for a sales conversation, the learning system should capture and assess that conversation’s outcome as a direct result of the learning.
Some businesses, like Capgemini, are already piloting this kind of skill proficiency assessment as a more accurate representation of business value.
How Capgemini Connected Skill Progression to Business
“I’m spending this amount of money on learning. Tell me what it’s delivering at the end.”
That’s what Estelle Maione, Global Head of Learning at Capgemini, was consistently hearing from executives. With over 300,000 employees and a company identity built around skills-based careers, Capgemini couldn’t just assume there was a link between learning and business performance.
“Since day one, we said that, as a company, we want to turn what we do with our skills into business impact,” Maione said.
Here’s what Capgemini business leaders wanted from L&D: personalized learning at scale, visible skill progression, and proof of project readiness.
A completion rate says nothing about whether a team is ready to staff a high-stakes client engagement. A course badge doesn’t tell a leader what the team can actually deliver.
So Capgemini looked for a different kind of skill signal: project delivery and performance.
When someone is actively working on client assignments at a given proficiency level, they are effectively demonstrating their capability through the work itself, rather than in a controlled training environment. Real work data becomes skill data. This is key because, as Maione put it, “If your data is not accurate, magic doesn’t happen.”
Capturing project-based data required an interconnected ecosystem, and one where people data and skill data work together to paint a comprehensive picture. At that point, business performance becomes the proof, and learning’s role is to show how it contributed over time.
“We are connecting with our operation system, which is where things happen because people are staffed on projects. That’s real life. That’s how people build their skills,” said Maione.
The result is a reporting model that can actually respond to the executive ask for measurable deliverables: When you spend X on learning, you get Y proficient employees who can deliver client outcomes.
“It’s really about how we put all our skills in one place and how we are making insights or translating the information we get into something that is valuable for business decisions, talent decisions for employees, and also to drive their skills development,” Maione said.
Even with a unified view of skill data, this remains a cross-functional effort. At Capgemini, the skill framework doesn’t reside with L&D. “It’s sitting with our operations teams because they have the most holistic vision of what’s needed,” said Maione.
For Maione, the principle is clear: Even when learning itself is invisibly embedded within daily work, the business outcome should always be clearly visible.
What the Next Generation of Learning Measurement Looks Like
Capgemini’s approach represents the leading edge of what’s possible today. But as technical capabilities evolve, so does what’s measurable. Taylor Blake, SVP of New Initiatives and Head of AI Labs at Degreed, has been prototyping what might come next for more personalized reporting.
Instead of a completion percentage, imagine a report that surfaces an individual’s confidence and performance lift over the course of a program, derived from how they actually engaged throughout. Instead of a self-reported survey metric, it would capture the unique trajectory of that individual’s learning path, including their starting point, gaps, and how the program adapted in flight.
In the AI Labs demo, Blake mapped two fictional employees with different skill trajectories, showing not only completions, but also variation in skill progress. This included confidence growth and demonstrable proficiency side by side.
This makes learning evidence more comparable to other forms of business evidence. Confidence growth becomes a quantifiable metric. Proficiency derived from real interactions can even reveal whether someone is more confident than they are capable—or the reverse. That’s a signal no completion report can give you.
How to Reimagine the Connection Between Learning and Business Value
Learning metrics need to live closer to business reality. Loosely inferring that completions signal workforce readiness doesn’t equate to measurable business value. It’s questionable reasoning at best. But thanks to emerging AI capabilities and a growing body of skill data, L&D can do more than just check boxes.
Now, when evaluating learning, HR and L&D need a real answer to the question: Did the workforce get better at something that matters to the business?
Getting there requires two things: skill data that’s clean and trustworthy, which means normalized taxonomies and consistent definitions across systems, and learning activity connected to the performance signals business leaders already use to evaluate workforce readiness.
Like Capgemini, the organizations that build this foundation now won’t just measure learning better. They’ll make faster, more confident decisions about their workforce, and they’ll do it in anticipation of business needs.
Let’s keep in touch.
Sign me up for the monthly newsletter with exclusive insights, upcoming events, and updates on Degreed solutions.