Skip to content
Sign in
← All notes

Why learning at work is moving from courses to continuous practice

Annual training assumed skills were stable enough to teach in a block and use for years. When the tools change inside a single role, learning has to live in the work instead.

Skills7 min read

The training model most organisations still run was designed for a world where capabilities were stable. Identify a gap, book a course, complete it, apply it for several years.

Every part of that assumes the thing being taught will still be current when it is needed. That assumption is now the weakest part of the design.

Why courses struggle now

A course has a long lead time. Somebody notices a gap, finds provision, schedules it, people attend, and it is delivered some months after the need appeared. That worked when the need was stable.

It also teaches to a general case, while the useful version of most capabilities is specific to your systems, your customers and your constraints. The general version transfers less than anybody hopes.

And it measures completion, which is not learning. Completion is the easiest thing to record and the least related to whether anybody can now do anything they could not do before.

The course model: a gap noticed, provision found, a session scheduled and delivered months later, recorded as completed. Every step assumes the requirement stayed still while it happened.
Completion is the easiest thing to record and the least related to capability.

What continuous practice means concretely

It does not mean abolishing training. It means moving the centre of gravity from events to the work, with the course as an occasional supplement rather than the mechanism.

  • Small, regular repetitions of the actual skill, with feedback close enough in time to be useful.
  • Learning inside real work rather than alongside it, so what is practised is the specific version that matters.
  • Feedback from somebody who can tell the difference between good and nearly good, which is the scarce ingredient and the one most schemes skip.
  • Evidence of capability rather than records of attendance, so somebody can see what they can now do.
Learning moved inside the work: small repetitions of the real skill, feedback close enough in time to be useful, and evidence of what somebody can now do rather than a record of what they attended.
Feedback from somebody who can tell good from nearly good is the scarce ingredient.

Why this is harder than it sounds

Continuous practice sounds obviously better and is much harder to run, which is why the course model persists despite everybody knowing its limits.

A course is easy to procure, easy to schedule, and easy to report on. Continuous practice needs somebody to design the repetitions, someone competent to give feedback, and a way to record capability that is not attendance. It is real ongoing work, spread across managers who are already busy.

That difficulty is exactly why it is worth doing. It is not a thing a competitor can buy in a quarter.

If you are doing this for yourself

You do not need your employer to run this well. The same structure works alone, and the only genuinely hard part is arranging feedback.

  • Pick one capability per quarter, not five. Depth compounds and breadth does not.
  • Practise inside real work, on something with a consequence.
  • Find one person who is better than you at it and will tell you the truth.
  • Keep evidence of what you can now do. It is the record that survives a job change.

Common questions

Why are traditional training courses less effective now?
Three reasons: their lead time is long relative to how fast requirements change, they teach a general version of a capability whose useful version is specific to your context, and they measure completion, which is not the same as being able to do something new.
What does continuous learning at work look like in practice?
Small regular repetitions of the real skill inside real work, with feedback from somebody who can tell good from nearly good, and evidence of capability rather than records of attendance. The course becomes a supplement rather than the mechanism.
How do you measure learning if not by course completion?
By what somebody can now do that they could not before, demonstrated on real work. That means capturing evidence: a decision they made and its outcome, a problem they handled unaided, a piece of work assessed against a standard.
Can employees do this without support from their employer?
Mostly, yes. One capability per quarter, practised inside real work, with evidence kept. The genuinely hard part alone is feedback, so the highest-leverage move is finding one person who is better than you and will tell you the truth.

PDP Quest exists because of the problem underneath all of these: when output stops indicating capability, you need another way to know who can actually do the work.

See how verification works →