Peer learning: the cheapest form, and the worst organised

Echipa HR 365 · reviewed 2026-09-13 · 5 min read

Most of what someone knows about their job they learned from a colleague, while working. It is the cheapest and most effective form of training a company has — and the only one that appears in no budget, has no owner, and is not organised at all.

Why it works better than a course

Why it does not happen more often

1. It is not recognised as work. Whoever explains loses their own time and it shows up nowhere.
2. Nobody knows who knows what. Without a map of skills, you ask whoever is nearest.
3. There is no moment. Without a declared slot, “when you have time” means never.
4. It is assumed to happen by itself. Sometimes it does; usually only among people who already sit together.

The first cause is the strongest. In a company where time is measured by deliverables, an hour spent explaining to a colleague is an hour off your own results — and it does not appear in your review. The rational behaviour is not to explain, and people adopt it without ever saying so.

Three formats that work

From the cheapest to the most structured

1. The 20-minute session — someone shows the team one thing they do well; every two weeks, at the same time
2. Pairing on a task — two people do a piece of work together that one knows and the other does not; once, not as a permanent practice
3. Skill transfer — the four stages from the delegation article, applied deliberately to a function covered by a single person

The first has the best effort-to-effect ratio and needs only one decision: a fixed slot in the calendar where anyone can show something. Without the fixed time, it is cancelled at the first busy week and never resumes.

What is worth showing

Not presentations about big subjects. What works is small and concrete: how to do a routine operation faster, what someone learned from a mistake, how to handle a case that comes up often and that everybody handles differently.

The second row deserves explicit encouragement, because it does not happen on its own: in a culture where mistakes are not discussed, nobody volunteers to narrate one. The first time, the manager does it, with a mistake of their own — otherwise nobody follows.

How to make it count

• It appears in the review, as real work. Someone who trained two colleagues this year did a job, not a favour.
• It counts towards workload. An hour explaining is an hour subtracted from somewhere.
• It is named in public. “X showed us how they do Y” costs ten seconds and is the best form of recognition there is.

What if nobody volunteers to show anything?

Do not ask “who wants to”; ask one specific person, about a specific thing you know they do well. After two or three sessions that are visibly short and useful, others volunteer.

Should sessions be recorded, or should there be materials?

Not at the start — the requirement adds effort and stops people. If a session proves very useful, it gets repeated, and that is when it is worth turning into material.

Where to start

Put a twenty-minute slot in the calendar, once a fortnight, and ask the first person to show one concrete thing. Do not announce a programme — announce a session.

By the third session, check whether anyone applied anything. If they did, you have a habit; if not, the subjects were too general and are worth changing, not abandoning.

What does not transfer well between colleagues

Not everything is learned from colleagues. There are three categories where the method works badly and it is worth knowing them, so that you do not build a programme on something it cannot deliver.

The second row is the most dangerous, because it is invisible: a wrong practice passed from colleague to colleague becomes, in two years, “how we do it here” — and nobody can say any more where it started.