Day 1 · 17 August 2026

Lean Learning — 20% Content, 80% Practice

Only about 12% of employees apply what corporate training teaches them. Lean learning is the response: minimum effective content, immediate application, immediate feedback, repeat.

Lean Learning — 20% Content, 80% Practice
lean learning80/20micro learning

The idea and its source

Lean learning is the answer to the forgetting curve — and the framework comes from the same Steve Glaveski HBR article as the retention figures, modelled on Toyota's lean manufacturing: effort only where it is needed, waste cut, outcomes improved. His sharpest statistic: only about 12% of employees apply the skills from corporate training to their jobs. The cycle: learning the core of what you need to learn, applying it to real-world situations immediately, receiving immediate feedback and refining your understanding, then repeating the cycle. The workbook's definition: getting the right learning, to the right audience, at the right time, in the right quantity, to achieve perfect work flow — while minimising waste and being flexible to change. My margin notes say it plainer: you do not have much time — hands on, practical. Adults will not give you 40 spare hours; design for the minutes they actually have, as close to the work as possible.

How to apply it

From the slide, the application checklist. Think 80/20 — 20% content, 80% practice. Apply learning to real-world situations. Leverage guided learning that continues into live application. Personalise content to each employee's performance and learning style. Provide ongoing support. Activate peer learning — 55% of employees ask a colleague first when they need to know something. Use micro courses, bite-sized. And move from credits to outcomes: learn the right thing, at the right time, for the right reason, and retain it. The 55% figure deserves its own sentence: half of all workplace learning already flows through colleagues, unmanaged. Peer learning is not a nice-to-have channel to add — it is the dominant channel, waiting to be designed instead of left to luck.

Case Study

a lean onboarding system for a new property

Opening a new property means a batch of new hires must absorb dozens of SOPs in two weeks. Rebuilt lean: The classroom shrinks to half a day, covering only the irreversible things — safety, cash handling, data privacy — and even that half-day runs at 20% content, 80% drills. Everything else moves to the point of use: a laminated micro-card at each workstation, a two-minute video linked by QR code, then doing the task under observation. That is micro courses plus apply-immediately, literally standing where the work happens. Peer learning gets designed instead of assumed: every new hire is paired with a buddy from an existing property for two weeks — since 55% of people ask a colleague first anyway, we choose which colleague. Buddies give feedback with the three-part frame: specific praise, then improvements, then what to avoid. Ongoing support is a group chat where any SOP question gets answered within the hour, and the answers get folded back into the micro-cards — minimising waste, flexible to change. Same material, roughly the same total hours as the old three-day marathon. But the hours are redistributed from one heroic learning event into a spaced, social, hands-on system — from teaching to whistling.

Applying it same-day — a lean design for my own AI course

Lean learning says apply immediately, so I did — the same evening I redesigned the AI course I teach at DJC around today's frameworks, and pressed the design into a one-page poster (below). The spine is a repeatable lean learning cycle: learn (the minimum needed right now) → see it (trainer demonstrates) → do it (apply to a real problem immediately) → get feedback → refine (make version 2) → repeat (harder problem) → apply (in real work) → get results. The content is restructured as mission-based learning — every session delivers a real outcome the learner can use that day: turn a 30-minute task into a 5-minute task, produce a week of marketing content from one idea, build a first AI workflow. Two mechanics come straight from today's material: a 20:80 balance (trainer explains 20% of the time, learners produce 80% — the trainer is the guide, the learner is the producer) and the 3-2-1 retrieval to close every session, reconstructed at the start of the next one before notes come out. The promise writes itself once the design is lean: we don't just teach the tool — we help people solve real problems and get results today.

Click to view with notes

Reflections

Day 1 ends with the frameworks snapping together into one sentence: the classroom is only 10%, whatever it delivers decays on the forgetting curve, so design lean — minimum effective content, maximum immediate practice, spaced retrieval, and deliberately engineered learning from others and from the work itself. Tomorrow I want to connect this to how adults specifically learn.