From Information to Capability

The L&D model that built the modern enterprise is quietly breaking. Here’s what replaces it.

For thirty years, learning and development has been an information problem. Build the course. Assign it. Track completion. Move on.

The model worked when the bottleneck was access. People needed to know things they didn’t know, and the LMS was the cheapest way to get a fact in front of a face. Compliance modules. Product training. Onboarding decks. Click. Click. Done.

But the bottleneck moved. It moved years ago. We just didn’t have a tool that let us see it. The real bottleneck is capability.

Every L&D leader already knows the truth about their LMS: completion is not competence. A rep who finishes the certification module still can’t run the discovery call. Watching the leadership video doesn’t make a new manager ready to deliver a performance review without making it worse.

The gap between “they took the training” and “they can do the job” is where the industry has been stuck for a generation. And until very recently, the only way to close it was to put a human coach next to every learner. Expensive. Slow. Impossible at enterprise scale.

So we settled. We accepted that training delivers information, and that capability builds on the job, slowly, by accident, at huge cost.

Experiential learning is what closes the gap.

This isn’t a new idea. The 70-20-10 model has been around for forty years. The research has been clear for longer than that: people build capability by doing the thing, getting feedback, and doing it again.

The new part is that AI finally lets us do this at scale.

A new manager can practice the difficult conversation before they have it. A sales rep can pitch a real buyer persona, hear a real objection, and listen back to what they sounded like. A first-line leader can rehearse the layoff conversation in a private space, with feedback, before they ever sit across from a person.

This is what the next decade of L&D looks like. Practice. Feedback. Repetition. Running at enterprise scale for the first time ever, beyond what any video library or manual-summarizing chatbot could touch.

What this looks like in practice

The shift shows up in four places first.

Manager development. Performance reviews. Difficult conversations. The talks that determine whether someone stays or leaves. Today, most managers walk in cold and learn on the live person. Experiential learning lets them practice the conversation before it happens, with a coach that doesn’t run out of patience.

Snowflake’s enablement team was watching 162 direct managers each burn 7.5 hours a quarter grading pitch and roleplay submissions, more than 1,200 hours company-wide, just to keep certification consistent across regions. They handed that grading to Yoodli’s real-time, rubric-based feedback and got the hours back without softening the bar. The real surprise showed up in what reps did with the extra room to practice. “Our reps aren’t just completing an assignment,” said Nathan Irby, Principal Technical Platform Enablement Strategist at Snowflake. “They’re motivated to keep practicing to get sharper.” That’s the real signal. When practice helps, people keep doing it after the requirement disappears.

Onboarding and ramp. Ramping 3,000 sellers in weeks instead of quarters becomes a quarter-end outcome. Standardized practice and evaluation across roles, regions, and cohorts, run the same way for everyone.

Leadership development. Managers drive 70% of the variance in employee engagement. The training should look like it matters that much. Real practice on the conversations that move the number, not a Friday-afternoon video.

LAK Group built their whole coaching model on this idea: leaders rehearsing tough conversations and feedback moments with Yoodli before they walk into the room, a shift that drove a 14x return on their coaching program and cut live coaching hours by two thirds.

Certification. Your rubrics. Your personas. Your standards. Practiced consistently across the org instead of measured on a multiple-choice quiz that no one believes in.

The proof is in the numbers

This is a measurable shift, and the L&D teams who got here first have the data.

• Snowflake reclaimed 1,215 manager coaching hours per quarter and saved $700K annually with AI roleplays.
• Google Cloud certified 15,000+ GTM employees in 30 days, with 92% CSAT.
• Harness cut review time on training submissions by 75%, saving roughly $94,500 per sales enablement team member, per year.
• LAK Group measured 14x ROI on a leadership coaching program built on experiential practice.
• Accepted Egypt ran 75 learners through 9,300 AI roleplays in six months. 47% landed BPO jobs. Salaries 2.2x on average.

Different industries. Different problems. Same underlying pattern. The moment you let people practice the thing and get feedback at scale, capability moves.

Why this works now and didn’t before

The unlock is multi-persona roleplay. Until recently, AI practice meant one scripted character reacting the same way every time, useful for a first pass, but nothing like a real conversation. Now the same session can hold several personas at once: the skeptical VP, the direct report who gets defensive, the buyer who won’t commit, and the AI shifts between them the way a real room does. That’s what makes the practice transferable. A manager who has rehearsed a layoff conversation against a persona built to push back, go quiet, and get emotional is ready for the actual conversation, not just the easy version of it.

A question for L&D leaders

The shift here is from passive to active, from a course your people sit through to a practice they do.

If you measured your team’s capability the way a coach measures an athlete, by reps taken instead of hours watched, what would your number look like today? And what would it look like in six months if every person you serve had access to a coach that never got tired?

That’s the L&D function that’s coming. Fewer decks. Better reps.

Book 15 minutes with our team to see what experiential learning looks like running in your org. yoodli.ai/talk-to-sales