Psychological Safety
Psychological Safety
What is psychological safety?
- A climate where mistakes and dissent are safe to voice
- The foundation of team learning and innovation
- Safe to be candid, not soft
What psychological safety really changes is not the mood but the speed at which information moves. In a low-safety team, bad news such as defects, delivery delays, and customer complaints takes days to travel up from the bottom, and the meeting room stays silent while the real conversation happens in hallways and messaging apps afterward. By then the problem reaches the leader with its cleanup cost multiplied. In a team with safety in place, problems are reported while they are still small, and a remark like “I don’t think that approach will work” comes out before execution and stops a bad decision. That is why the results of this concept show up not in satisfaction scores but in time to report and the cost of rework.
Edmondson plotted psychological safety and performance standards on two axes and sorted teams into four zones. High standards without safety is the anxiety zone; safety with low standards is the comfort zone; both low is the apathy zone; and only when both are high does a team reach the learning zone. Another distinction that often gets confused in practice is the line between safety and trust. Trust is a person-to-person judgment about whether you believe a particular individual. Safety is a shared, collective sense that “it is OK to do this on this team,” so it varies widely by team leader even within the same company. That is why survey items such as “Do I face penalties for making mistakes?” and “Can I raise an awkward issue?” are useful only when broken down by team rather than averaged across the company.
The most common misconception is equating safety with “a team that avoids saying unpleasant things.” A team that postpones performance feedback and covers for low performers is not safe; it is a team that has lost accountability, and its most conscientious members burn out and leave first. The second is treating it as something a leader can settle with a single sentence. If a team leader says “feel free to speak up” but has already laid out their own conclusion, there is no room for dissent in that meeting. The third is the price of speaking up. When someone who raises a problem is told “then you take care of it,” members learn exactly that the cost of speaking up outweighs the cost of staying silent, and the meetings go quiet from the next one on.
Seungbin JeonAn instructor who draws on psychology expertise to support communication and leadership growth in organizations
John KimInstructor John Kim, author of B2B Marketing Design and CEO of Nanum ConsultingAmy Edmondson of Harvard Business School defined it in a 1999 paper as “a shared belief that the team is safe for interpersonal risk taking.” It became widely known when Google’s Project Aristotle (2012–2016) named it the top factor in team effectiveness.
Consider a hypothetical case: the production division of Company A, a Korean auto parts maker with 240 employees and KRW 62 billion (≈ US$46,386,400) in annual revenue. Within six months it received 11 customer claims, and on review, most involved warning signs the shop floor already knew about that never made it upward. Rather than demanding to know why nobody spoke up sooner, management set out to find where the speaking up was getting blocked.
- Every Monday, the plant manager opened the first 10 minutes of the quality meeting by bringing up his own misjudgments. His first words were, “Delaying the decision to expand our equipment by two months was my mistake.”
- A quality team leader removed the “person responsible” field from the defect report form and added a “when did you first notice something wrong?” field. Once the form asked about timing instead of blame, reports came in noticeably faster.
- At a meeting, the Production Team 1 foreman said, “This fixture has been a problem for three years and nobody speaks up.” The plant manager did not push back; he only asked, “Since when?”
- HR ran a 7-question anonymous survey at the team level every quarter. The first-round average for the production division was 3.1, but team scores ranged from 2.4 to 4.0.
- They ended the practice of making whoever raised a problem take on the work too, and separated the two: improvement tasks that were raised are now handled by a three-person process improvement task force.
- Six months later, early reports of process anomalies rose from 4 to 19 per month, rework costs fell from KRW 23 million (≈ US$17,200) to KRW 14 million (≈ US$10,500) per month, and the survey average reached 3.9.