Workplace Violence: 7 Controls Before Reporting Fails

Workplace violence rarely starts with a dramatic event; it usually starts with threats, intimidation, client aggression, domestic spillover, or retaliation that the organization normalized as friction. This guide gives EHS and HR leaders seven controls that make those signals visible before reporting becomes a late administrative ritual. When violence becomes a critical incident, the response also needs post-traumatic stress controls for responders.

Why workplace violence is a work-design risk, not only an HR case

Workplace violence prevention fails when the company treats aggression as a personality problem after the event instead of a predictable exposure created by tasks, staffing, access, authority, customer interaction, and weak escalation paths. ILO C190, adopted in 2019, recognizes violence and harassment as a world-of-work issue, which means the boundary is wider than the factory gate or office desk.

As Andreza Araújo argues in Safety Culture: From Theory to Practice, culture is not what leaders declare during campaigns. Culture is what workers can do, say, report, and challenge when pressure is present. That distinction matters because violence indicators often sit beside workload risk indicators, silence, and poor supervision long before a formal complaint is filed.

The practical audience for this article is the EHS manager who must work with HR, security, legal, operations, and supervisors without letting the topic disappear into policy language. The reader should leave with a decision: map violence exposure by task and control quality, not by the number of complaints received.

1. Map exposure by task, not by department

Workplace violence exposure becomes visible when the assessment starts with the task, because the same department can contain low-risk administrative work and high-risk contact with angry customers, isolated night work, cash handling, disciplinary meetings, or visits to unstable field locations.

The common mistake is to ask whether a department has a history of violence. That question rewards underreporting. Across 25+ years leading EHS at multinationals, Andreza Araújo has observed that exposure without incident history can still be serious when the task has public contact, lone work, authority conflict, or time pressure.

Start with a task inventory. Flag work where employees deny service, enforce rules, carry valuables, enter private sites, supervise conflict, handle complaints, or work without immediate backup. The control owner should be named for each exposure, because a risk register with no owner is only a document.

2. Treat threats and intimidation as precursor events

Threats, intimidation, stalking, aggressive gestures, and repeated hostile messages are precursor events, not soft complaints. Heinrich and Bird's accident-pyramid logic applies here because serious harm is usually preceded by weaker signals whose meaning is missed or discounted.

What most organizations miss is that a violence precursor does not need physical contact to deserve EHS attention. When supervisors wait for injury before acting, they train the workforce to stay silent until the situation becomes impossible to deny. That pattern is close to the reporting failure described in bad news in safety, where leaders say they want truth but punish the messenger through delay, disbelief, or exposure.

Create a simple precursor taxonomy with four levels: concern, threat, attempted harm, and harm. Each level should trigger a response time, a responsible role, and a decision about separation, security, investigation, or support. If the taxonomy is too legalistic for supervisors to use, the field will ignore it.

3. Separate reporting from investigation

Reporting workplace violence should be easier than proving workplace violence, because the first action is to surface risk, not to win an evidentiary argument. A report is an early warning; an investigation is the structured process that tests facts, protects people, and decides corrective action.

In more than 250 cultural transformation projects, Andreza Araújo has seen that employees often stay silent when the same manager who receives the report also controls schedules, overtime, evaluations, and discipline. The reporting path then becomes contaminated by fear, especially where the alleged aggressor has status or influence.

Use at least two reporting channels, one through the line and one independent of the line. Define immediate protection steps that do not require a completed investigation, such as shift separation, escort, access changes, meeting controls, or removal from public-facing duties when risk is credible.

4. Protect confidentiality without hiding patterns

Confidentiality protects people from retaliation, but excessive secrecy can hide repeated exposure across teams, locations, and shifts. The system has to protect identities while still allowing EHS, HR, security, and senior leadership to see pattern data.

This is where many programs fail. They promise anonymity, then collect reports in a way that makes follow-up impossible, or they protect case details so tightly that no one sees the trend. The better design is role-based access, where case facts stay limited and aggregated signals reach leaders through a dashboard.

Track repeat locations, repeat aggressors, task type, time of day, trigger, control failure, and response time. If a site has no reports for a year but has high turnover, tense customer interactions, and supervisors who discourage escalation, the absence of records should be treated as a signal, not proof of health.

5. Build supervisor scripts for the first ten minutes

The first ten minutes after a threat decide whether the employee feels protected or exposed. Supervisors need a script because improvisation under pressure often turns into disbelief, minimization, or a promise the supervisor has no authority to keep.

As Andreza Araújo writes in Make The Difference: Be a Leader in Health & Safety, operational leaders shape safety through concrete behaviors that workers can see. For workplace violence, visible felt leadership is not a speech about respect; it is the supervisor asking what happened, separating people when needed, preserving information, and escalating within the defined time window.

The script should contain five moves: ensure immediate safety, listen without cross-examining, record the exact words or behavior, explain the next step, and confirm a follow-up time. Supervisors should avoid promising outcomes, diagnosing motives, or asking why the employee did not report sooner.

6. Connect violence controls to ISO 45003 and ISO 45001

ISO 45003 gives organizations a practical language for psychosocial hazards, while ISO 45001 requires control of occupational health and safety risks through planning, operation, performance evaluation, and improvement. Workplace violence belongs inside that management system, not beside it as a separate poster campaign.

A useful control plan links violence exposure to the hierarchy of controls. Elimination can mean redesigning a service process that forces employees to deny benefits alone; engineering can mean access barriers or panic alarms; administrative controls can mean staffing, scripts, escalation, and case review; PPE is rarely the decisive barrier in this risk family.

That hierarchy is the reason a violence program should connect with psychosocial risk controls and not only with employee conduct rules. The target is not only discipline after aggression. The target is designing work so aggression is less likely, less hidden, and less able to escalate.

7. Audit response time and corrective action quality

Workplace violence metrics should measure whether the organization acts fast enough and learns well enough, because complaint counts alone are distorted by fear, trust, and reporting access. A low report count can mean low exposure, or it can mean people stopped trying.

Use response indicators that leaders can audit: time from report to first contact, time from report to interim protection, percentage of cases with task-level corrective action, repeat exposure after closure, and employee confirmation that the agreed protection was implemented. These indicators are stronger than a yearly count because they test system performance.

During Andreza Araújo's PepsiCo South America tenure, where the accident ratio fell 50% in six months, the lesson was not that one metric solved safety. The lesson was that leadership attention, operating discipline, and fast correction change results when the system measures the right work.

Each month without response-time auditing allows weak signals to become informal folklore, while formal dashboards continue showing a silence that leadership may mistake for control.

Comparison: policy-only program vs control-based program

Conclusion: violence prevention needs operating discipline

Workplace violence prevention becomes credible when leaders stop treating reports as isolated HR files and start managing exposure, precursors, response time, and work design with the same discipline used for SIF controls.

For leaders who need to build that discipline across culture, psychosocial risk, and operational safety, Andreza Araújo's Safety School and ACS Global Ventures consulting work help organizations move from policy compliance to real conversations, real controls, and work that protects people. Safety is about coming home.