Andrea Hernandez on Why Psychological Safety Starts in the Smallest Team

Psychological safety is often announced at the largest possible scale. The company publishes a message, adds speaking-up language to a leadership program, and asks thousands of employees to share concerns. Yet the real test happens in a much smaller room, on the shift, in the control room, or during the handover where a person decides whether it is safe to say, “I am not convinced.”

Andrea Hernandez, Siemens Global Health and Safety Vice President, made that point in Episode 8 of the Headline Podcast. Her argument is operational rather than ceremonial. People do not begin by trusting an enterprise statement. They begin by reading the response of the team that shares their work, their supervisor, and the colleague who will be beside them when the task becomes difficult.

Psychological safety starts in the smallest team because that is where workers experience the consequences of asking questions, reporting weak signals, and challenging an unsafe assumption.

Why the smallest team is the first safety test

A workforce can receive the same values statement and still experience very different levels of safety. One team may treat a late stop as responsible judgment, while another treats it as a failure of commitment. One supervisor may ask what information is missing, while another asks who caused the delay. The policy is identical. The operating climate is not.

Amy Edmondson's research on psychological safety helps explain why this local response matters. People decide whether to speak after reading social risk, not after memorizing a definition. They notice who is interrupted, who is mocked, whose concern receives follow-up, and whether the person who brings bad news is welcomed into the next conversation.

That reading becomes especially important in safety work because uncertainty is normal. Conditions change, controls degrade, and the person closest to the task often sees the first indication that the plan no longer fits. If the team punishes uncertainty, the organization loses information before a formal incident occurs.

The first question is not whether the company has a speaking-up policy. It is whether the nearest supervisor makes a difficult question easier to ask.

What a global safety strategy misses without local context

Andrea Hernandez's conversation also places context before slogans. A global strategy can define intent, resources, and minimum expectations, but it cannot predict every local pressure that shapes a decision. Staffing, production demand, language, contractor relationships, equipment condition, and the supervisor's authority all influence what people believe they can say.

This does not mean that each site should invent its own standard. It means that leadership must examine how a standard is interpreted at the point of work. A rule that appears clear at headquarters may become ambiguous during a night shift, a maintenance interruption, or a handover in which two teams believe the other owns the decision.

Andreza Araujo's work on the gap between declared commitment and operated conditions provides a useful lens here. A safety value becomes credible when the organization funds it, gives people permission to act on it, and accepts the operational consequence of the safer decision. A poster cannot provide that permission.

Global consistency should define the safety promise, while local leaders must prove that the promise survives schedule pressure, uncertainty, and inconvenient information.

How silence is produced during ordinary work

Silence rarely begins with an explicit instruction to hide a concern. It develops through small exchanges that teach the team what carries a cost. A supervisor rolls their eyes at a question, a manager praises the shift for avoiding reports, or a worker who stops a job is later excluded from the planning discussion. None of these acts needs to be described as retaliation for the lesson to be understood.

People then adjust their behavior. They wait until a concern is certain, soften the language, tell a colleague privately, or solve the problem alone. That delay matters because safety information loses value when it arrives after the exposure has moved, the equipment has been reset, or the decision has become difficult to reverse.

The silence can also look like agreement. A meeting with no questions may indicate clarity, but it may indicate that the group has learned the leader prefers speed over challenge. Leaders should not infer trust from quietness when the decision is complex and the consequences are serious.

Silence is a safety signal when people wait to speak until the evidence is undeniable, because the team has made early uncertainty socially expensive.

What supervisors should ask before work begins

The most useful test is built into the work rather than added as a survey. Before a critical task starts, the supervisor can ask which assumption is least certain, what has changed since the plan was approved, and what would make the team stop. These questions do more than collect opinions. They define dissent as part of competent preparation.

The supervisor should also ask who has authority to pause the task and what response the person should expect after doing so. Authority that exists only in a policy is fragile. Workers need to know whether the pause will be investigated fairly, whether the schedule will be adjusted, and whether the control will be restored before the task resumes.

A good conversation does not require a long meeting. It requires a real response. If a worker identifies a missing isolation, an unclear interface, or a change in weather, the team should show how the information changes the plan. Otherwise, the question becomes a ritual that trains people not to bother next time.

Before work begins, ask what may fail, what has changed, who can stop the task, and what evidence will show that the concern was addressed.

How to respond when someone challenges the plan

The first response carries more weight than the later campaign. Leaders should acknowledge the concern before defending the plan, clarify what is known, and invite the person to explain the observation in operational terms. This sequence protects the evidence from becoming a debate about confidence or loyalty.

There will be cases in which the concern does not require a stop. That outcome does not make the report unhelpful. A team can examine the evidence, decide that the control remains adequate, and explain the reasoning without humiliating the person who raised the question. The quality of the response is measured by the clarity of the decision and the respect shown during the review.

When leaders respond defensively, the team learns that speaking up requires a courtroom standard. That standard is unrealistic for early signals, because early signals are often incomplete by definition. A mature response therefore separates uncertainty from error and investigation from accusation.

A challenge strengthens safety when leaders examine the observation before judging the person, then explain the decision and the evidence that supports it.

How well-being belongs inside safety leadership

Andrea Hernandez's Siemens perspective connects well-being with company strategy rather than treating it as a side program. That connection matters because work design affects both physical exposure and mental strain. A schedule that leaves no recovery time, a monitoring system that creates constant interruption, or a target that makes stopping feel impossible can weaken judgment before anyone describes the situation as a safety problem.

Leaders should therefore examine the conditions that make people less willing or less able to speak. Fatigue, isolation, workload, uncertainty, and conflicting priorities change the social risk of raising a concern. The worker may understand the hazard and still decide that speaking will create a larger problem for the team.

ISO 45003:2021 gives organizations a structured way to consider psychosocial risks within an occupational health and safety system. The practical value is not the label. It is the reminder that leadership, work organization, communication, and relationships can create conditions that affect health and safe performance.

Well-being becomes a safety issue when workload, pressure, or work design makes people less able to notice, discuss, or act on changing risk.

How leaders can verify that psychological safety is real

Leaders should verify behavior, not only sentiment. A team may score well on a survey while still withholding concerns in the moments that matter. Verification should compare what people say in a scheduled discussion with what happens during a changing task, after a near miss, or when a target is at risk.

Use a small set of evidence points. Review who raised concerns, how quickly the response occurred, whether the plan changed, whether the person received feedback, and whether similar concerns appeared again. The purpose is not to rank supervisors by report volume. A high number may indicate trust, exposure, or both. The useful question is whether the organization made the information actionable.

Leaders can also conduct short listening sessions without the direct supervisor present, then compare the themes with field observations and incident reviews. If workers describe a strong speaking-up climate but the evidence shows repeated late escalation, the difference deserves attention rather than a favorable average.

Psychological safety is credible when early concerns lead to visible decisions, respectful follow-up, and controls that remain usable after the meeting ends.

Recommendation: make the next small conversation count

The next practical step is not another enterprise slogan. Choose one team that performs consequential work and observe how it handles uncertainty before, during, and after a task. Ask the team what makes a concern difficult to raise, then listen for the specific response that would reduce that difficulty.

Give the supervisor authority to change the plan and require a short record of what changed, who owns the follow-up, and when the team will verify the control. If the concern is rejected, record the reasoning. Transparent disagreement is safer than silent compliance because it keeps the decision visible to the people who must live with it.

Andrea Hernandez's central lesson is practical. Psychological safety is not created by the size of the audience that hears the message. It is created by the quality of the response in the smallest group that shares the work. When that group can question an assumption without losing status, leaders receive better evidence before a weak signal becomes an incident.

For a related test, read the five signals that silence critical information and the guide to keeping technical dissent usable under pressure. The Headline discussion on trust as a safety-culture test adds another leadership lens. Together, these pieces point to the same decision: make it easier for the person closest to the work to tell the truth early.