We have spent the first two parts of this conversation looking at the people behind Trust & Safety work. First, we asked a question that is often overlooked: Who protects the people protecting the internet? Then we looked more closely at what the work can do to the people doing it. Stress, burnout, secondary traumatic stress, vicarious trauma and compassion fatigue are not interchangeable experiences, but they can all become part of the human reality of work that involves repeated exposure to difficult, disturbing or traumatic material.
We close this blog series by looking at “what should organisations actually do differently?”
It is not enough to recognise that Trust & Safety practitioners can be affected by their work. Recognition without structural change can quickly become another form of organisational rhetoric. An organisation can offer a wellness session, circulate a mindfulness resource, provide an Employee Assistance Programme and remind employees to take care of themselves.
But if the same organisation continues to design work around excessive exposure, relentless queues, unrealistic performance expectations, inadequate recovery and a culture where asking for help is quietly interpreted as weakness, then the underlying problem has not been addressed.
Wellness cannot become a substitute for fixing the workplace.
Resilience should not mean endurance. There is a tendency to celebrate resilience in Trust & Safety. The work requires people who can make difficult decisions, remain focused under pressure, deal with disturbing information and continue functioning when the volume of work is high. However, it is important we distinguish between resilience and endurance.
Resilience is not the ability to absorb unlimited pressure without showing signs of harm.
If an organisation continually increases a person's workload because that person has demonstrated that they can handle it, it may eventually turn competence into a vulnerability. The person who never complains receives the most difficult cases. The person who consistently clears the queue receives more work. The person who can manage high-severity investigations becomes the person most frequently exposed to them.
A resilient practitioner should not have to compensate for an unresilient system.
Resilience needs to become a systems question. Instead of asking only whether employees have access to counselling, resilience should be discussed when teams set staffing levels, design workflows, establish targets and allocate high-severity cases.
Organizations should ask themselves: “what changes can we make to the work before asking people to become better at coping with it?”
Start by measuring exposure, not just output
Trust & Safety organisations are generally very good at measuring work. They know how many cases are in a queue, how quickly they are being processed, how accurate decisions are and whether service-level expectations are being met.
What is often less visible is what kind of work people are being exposed to in order to produce those numbers. An organisation should therefore begin tracking exposure alongside productivity. That does not mean turning a person's psychological wellbeing into another performance metric. It means understanding where risk is concentrated.
How frequently is someone assigned high-severity content? How long are they spending on it? Are the same people consistently handling the most disturbing cases? Are certain teams carrying significantly more exposure than others? Are employees working beyond scheduled hours to keep up with demand?
These questions can be incorporated into existing operational reviews rather than creating an entirely separate wellbeing bureaucracy. For example, a weekly team review could look not only at queue volume and SLA performance, but also at high-severity case volume, staffing capacity, overtime, exposure concentration and recovery capacity.
If the numbers repeatedly show that a small group is carrying disproportionate exposure, that should trigger a management conversation.
Design exposure into the workflow
Some exposure is unavoidable. A child safety investigator cannot investigate child exploitation without encountering evidence of exploitation. A content moderator cannot enforce policies without seeing some of the content those policies prohibit. We cannot achieve zero exposure, however, we should look at whether it is necessary exposure rather than unnecessary exposure.
High-severity cases can be routed through defined escalation pathways rather than requiring large numbers of practitioners to independently review the same material. Where appropriate, technology can be used to classify, filter, blur or otherwise reduce unnecessary exposure before human review.
Teams can also introduce structured rotation so that the same practitioner is not continuously assigned the most psychologically demanding work. Importantly, rotation should not simply mean moving people from one queue to another. A meaningful rotation model might deliberately alternate high-intensity work with lower-intensity tasks, administrative work, quality review, training, policy work or other activities that require different forms of attention.
Put recovery into the schedule
Recovery should not depend on whether an employee is brave enough to ask for it. If breaks are regularly cancelled because queues are high, then the organisation does not really have a break policy, it has a theoretical break policy.
Organisations can make recovery more concrete by building it into workforce planning and scheduling. That could mean protected breaks that are not routinely sacrificed to meet queue targets, transition time after high-severity investigations and ensuring that shifts involving particularly demanding work are followed by appropriate recovery time rather than immediately placing the same practitioner into another high-intensity workload.
Teams can also establish a simple operational rule: when demand exceeds sustainable capacity, the response is to escalate capacity, not quietly transfer the excess burden to the people already working. That may mean activating additional staffing, reprioritising queues, adjusting service expectations or temporarily changing workflow.
Without that mechanism, “resilience” often becomes a euphemism for overtime.
Make managers part of the protection system
Managers should not be expected to diagnose trauma, burnout or mental health conditions. They should be equipped to recognise when the way work is being performed is becoming unsustainable. That requires more than generic management training.
Managers working with Trust & Safety teams should understand the nature of repeated exposure, know how to have a supportive conversation, understand available referral pathways and know what workplace adjustments they can actually make. A manager should be able to respond when someone says, “I am struggling with the volume of high-severity cases,” without immediately turning the conversation into a performance discussion.
More importantly, managers should have actual authority to change something.
There is little value in training a manager to identify workload-related stress if the manager has no ability to adjust schedules, redistribute cases, escalate staffing concerns or temporarily modify responsibilities.
Supportive management therefore needs two components: capability and authority.
Make help-seeking professionally safe
An Employee Assistance Programme is only useful if employees trust it. A counselling service is only useful if people believe accessing it will not quietly damage their career. A wellbeing policy is only meaningful if employees feel safe using it. This is especially important in high-performance environments where people may fear being seen as unable to handle difficult work.
Organisations should therefore make confidentiality boundaries clear, explain how wellbeing information is handled and separate support-seeking from routine performance evaluation wherever possible. Leaders can also change the cultural signal by normalising support without making disclosure compulsory.
“If the work is becoming difficult to sustain, there are safe ways to ask for support and ways we can change the work.”
Give practitioners a formal way to influence the work
People doing Trust & Safety work often know exactly where the system is creating unnecessary pressure. They know which queues consistently become impossible, which policies generate repeated escalations, which workflows expose them to material they could have been shielded from and when staffing levels do not match demand.
Organisations should create formal mechanisms for that knowledge to reach decision-makers.
This could be a monthly practitioner-led risk review, structured team debriefs, anonymous workflow feedback or a recurring review of high-severity exposure. The feedback should not be allowed to disappear into surveys. If practitioners repeatedly identify a problem, someone should be responsible for reviewing it, responding to it and communicating what will happen next.
“What changed because people told us?”
Change the conversation around performance
Trust & Safety requires performance measurement. However, organisations should be careful about creating performance systems where the only visible indicators are output and speed.
If an employee consistently works overtime to maintain their numbers, the system may record high performance while missing a warning sign. If a team meets every SLA by routinely sacrificing breaks, the organisation may celebrate operational success while quietly accumulating human risk.
This is where sustainable performance becomes important. Managers should be encouraged to look at performance alongside workload, staffing, overtime, turnover, absence and exposure patterns.
Are we achieving this target because the system is working well, or because people are absorbing the cost?
Create a clear response when risk increases
Organisations also need to decide in advance what happens when working conditions become unsustainable. Without a predefined response, every manager is left to improvise.
A Trust & Safety team could establish escalation thresholds for situations such as sustained excessive queue volumes, significant understaffing, unusually high concentrations of severe cases or persistent overtime. Once a threshold is reached, there should be an agreed response.
That might involve bringing in additional capacity, redistributing work, reducing exposure, adjusting priorities, temporarily changing service-level expectations or increasing management oversight.
The important part is that the burden does not simply move downward. When the system reaches capacity, the organisation should have a mechanism for absorbing the problem rather than expecting practitioners to absorb it themselves.
Treat wellbeing as an operational risk
Perhaps the biggest change organisations can make is to stop treating practitioner wellbeing as an optional programme sitting beside Trust & Safety operations. It should be part of how operational risk is understood.
The same way an organisation asks whether a new workflow creates privacy, security, legal or compliance risks, it should also ask whether it creates significant psychosocial risks for the people operating it. Before launching a new moderation queue, investigation process or performance target, organisations can ask:
How much exposure will this create? Who will carry it? How frequently? What recovery is available? What happens if demand increases? What controls exist if the workload becomes unsustainable?
These questions do not require an organisation to predict exactly who will experience burnout or trauma.
They require it to design for foreseeable risk rather than waiting for individual harm to become visible.
From resilient practitioners to resilient systems
The future of Trust & Safety cannot depend on finding people who are exceptionally good at absorbing difficult work. It should depend on designing systems that make difficult work sustainable.
That means:
Measuring exposure alongside output.
Designing workflows that minimise unnecessary exposure
Building recovery into schedules rather than treating it as a personal responsibility.
training managers and giving them the authority to act.
Making professional support genuinely safe to access.
Giving practitioners a meaningful voice in how the work is designed.
Recognising when operational targets are being achieved by quietly transferring the cost to employees.
Creating mechanisms that allow organisations to respond when demand exceeds sustainable capacity.
None of this eliminates the difficult realities of Trust & Safety work, but the people doing that work are not infrastructure. They are human beings.
And a system that depends on them remaining unaffected by repeated exposure to human suffering is not demonstrating resilience. It is relying on endurance.
The real measure of a resilient Trust & Safety organisation is therefore not how much pressure its practitioners can survive. It is how deliberately the organisation designs the work so that people can do it well, recover from it and remain well enough to keep doing it.
Because protecting the internet should never require us to sacrifice the people protecting it.
That is what it means to move from resilient practitioners to resilient systems.
About the Author
Athena Mwarwakamori Morgan is the Founder of Mindful Clicks Africa, where she advises schools, organisations, and policymakers on digital governance, child online safety, data protection, and responsible technology use. She has worked across Africa on child protection, technology policy, and safeguarding, helping institutions create safer and more resilient digital environments.