Learning isn't a perk.
It's infrastructure.

The skills half-life problem

Here's the uncomfortable truth: the shelf life of what your people know keeps getting shorter. A technical skill used to be good for a decade — now its half-life is closer to five years, and shrinking. Organizations that map this with a skills taxonomyA living inventory of the skills your teams have versus the skills the work demands — the foundation of any modern capability strategy. can see the decay coming. Organizations that don't find out the hard way: quietly, then all at once.

That's why deliberate upskilling and reskilling pathwaysUpskilling deepens ability in someone's current role; reskilling prepares them for a new one. Both cost less than replacing people from outside. stopped being a nice-to-have. They're maintenance — for your most expensive asset.

The cost of doing nothing

Everyone can see the L&D line in the budget. Almost nobody sees the bill for skipping it — because it hides in places finance doesn't label: people quitting, new hires taking months to ramp, mistakes that get quietly fixed and never counted.

Here's a number worth keeping: replacing one employee typically runs one-half to two times their annual salary once you add up recruiting, lost productivity, and the institutional knowledge that walks out the door with them.

Try it: plug in numbers that look like your organization, and watch what attrition actually costs.

Learning science ≠ content dumping

Most training doesn't fail because it was taught badly. It fails because brains are built to forget. Ebbinghaus proved it over a century ago: without reinforcement, most new information is gone within days. That's the forgetting curve, and every one-and-done compliance module runs straight into it.

The fix isn't longer courses. It's spaced retrieval practiceBringing material back at expanding intervals and making learners pull it from memory instead of re-reading it. One of the most robust effects in cognitive science. — bringing the material back at growing intervals and making people pull it from memory instead of re-reading it. It's also why scaffolded, scenario-based designEach interaction builds on the last, with realistic decisions instead of passive slides — managing cognitive load while maximizing transfer to the real job. beats a deck of slides: it makes learners use the knowledge while it's still warm.

Try it: drag the slider and watch what a few well-timed reviews do to 30-day retention.

Measuring what matters

"Did they like it?" is a fine question — it's just the smallest one you can ask. The Kirkpatrick model gives you four levels of proof, from smile-sheet surveys all the way up to business results.

Modern teams go further: they stream every meaningful click and decision as xAPI statements into an LRSA Learning Record Store collects Experience API (xAPI) statements — actor/verb/object records of everything a learner does. Far richer than the completions-and-scores world of legacy SCORM.. In fact, this course is doing it to you right now — check the log in the corner. Then click through each level of the pyramid.

You're the director

Your company just rolled out a big new enterprise platform, and it's going sideways. Error rates are climbing, the help desk is drowning, and the executive team hands you two weeks and a modest budget. You're the Director of L&D. What's your move?

Knowledge check

Nicely done.

You just moved through gated navigation, spaced concepts, a branching scenario, and embedded performance support — the same architecture I build for client programs. If a course like this would land with your team, let's talk.