Learning organisations turn experience into better decisions

The elements of a learning system

Shared goals clarify which outcomes matter. Psychological safety makes it more possible to ask, admit uncertainty or report an error. Feedback loops bring information back quickly enough to change work. Communities of practice allow tacit knowledge to move through relationships; documentation preserves decisions and evidence for later use.

Experimentation should be proportionate to risk. Reflection turns an event into an explanation and a next action. Measurement then asks whether behaviour or performance changed. No single book owns this combination, and each practice needs adaptation to power, regulation, safety and the work itself.

Choose the routine that addresses the failure

Edmondson’s 1999 field study examined 51 teams in one manufacturing company. It found an association between team psychological safety and learning behaviour. That supports taking speaking-up conditions seriously; it does not establish that introducing a meeting will cause the same performance gain in every workplace.

If colleagues do not know how to complete a handover, I would prioritise a worked example and observed practice. If they know the process but cannot find the latest instruction, I would fix the knowledge source and its ownership. If people see problems but fear reporting them, another training module misses the central issue. Choose the intervention after diagnosing the failure.

A practical maturity table

This table is an original discussion framework, not a validated organisational maturity scale. Use it to identify a missing routine rather than assign a scientific score to a team.

Activity is not outcome

Course completions, login counts and content views describe activity. They do not establish changed behaviour or value. Start from the performance question. If the goal is safer handovers, examine handover quality and incidents. If the goal is faster onboarding, examine time to competent independent work while watching quality and support burden.

Measure the change without overstating the result

Suppose a team records missing information in 8 of 40 handovers before a change and 3 of 40 afterwards. The recorded rate falls from 20% to 7.5%, a reduction of 12.5 percentage points. These are invented numbers showing how to report a rate, not results from a client or a study.

Keep the definition of missing information and the observation period comparable. Record workload and case complexity, and watch for people reporting fewer problems because scrutiny has increased. A before-and-after difference is a signal to investigate, not proof of causation. For consequential decisions, involve the people responsible for quality and safety in choosing the evaluation.

Technology supports the system; it is not the system

An LMS can organise enrolment, access, records and reporting. A knowledge base can make decisions findable. Neither creates trust, feedback or good practice by itself. Choose tools after defining the learning loop and governance. For a separate, commercial view of procurement, see current LMS options for workplace learning and training. That comparison is published under the same ownership and may contain affiliate links; this guide does not rank products.

Common pitfalls

• Treating training volume as evidence of capability.
• Collecting lessons without assigning a decision or owner.
• Inviting candour while punishing the person who raises risk.
• Running low-value experiments where a structural fix is already known.
• Measuring easy activity while ignoring the affected people and outcomes.