Most of the time, we're designing software for people who are mildly inconvenienced by a clunky interface. They'll sigh, click the wrong button, maybe complain in a meeting. The stakes are low. But when you're building case management tools for organisations that support survivors of domestic violence, the stakes are not low. The user might be in danger. The data might put someone at risk. The design decisions you make carry real weight.
What You Need to Know
- Designing for crisis contexts requires fundamentally different UX thinking than standard enterprise software
- Users in distress have reduced cognitive capacity. Every unnecessary step is a barrier
- Data security isn't a feature. It's a safety requirement. Breaches can put people in physical danger
- Cultural context determines whether a system will be used or abandoned
- "Move fast and break things" is the wrong philosophy when breaking things means real harm
The Context
We've been working on a case management system for an organisation in the Pacific that supports survivors of domestic violence. I'm keeping the details deliberately vague because the safety of the people involved matters more than a case study. What I can share is what we learned about designing for contexts where the user is under genuine duress.
This wasn't theoretical. We sat with case workers. We heard what their days look like. We understood, as much as outsiders can, what it means to document someone's worst experiences while also trying to help them find safety.
Cognitive Load Is Not an Abstraction
In standard UX work, we talk about cognitive load as a design principle. Reduce the number of choices. Simplify the layout. Use clear labels. It's good practice for any interface.
In crisis contexts, cognitive load is a survival constraint. Case workers using this system are often dealing with their own emotional responses to what they're hearing. They might be working with a client who is frightened, injured, or in immediate danger. They don't have the mental bandwidth for a complex interface.
50%
reduction in cognitive capacity under acute stress, affecting decision-making and information processing
Source: American Psychological Association, Stress Effects on Cognition, 2020
Every unnecessary field, every ambiguous label, every extra click is a real cost. We stripped our forms back to the absolute minimum required for case documentation. We used plain language, not clinical terminology. We made the most common actions the most obvious ones.
When someone is documenting a crisis, the software should disappear. If they're thinking about the interface, we've failed.
Rainui Teihotua
Chief Creative Officer
Data Security Is Physical Safety
In most enterprise contexts, a data breach means embarrassment and regulatory fines. In domestic violence case management, a data breach can mean someone gets found. That changes everything about how you think about access controls, audit logs, and data storage.
We designed the system with the assumption that someone with bad intentions might try to access it. Not a hacker on the other side of the world, but someone in the same community, maybe in the same household, who wants to know where their partner went. Role-based access isn't just a nice-to-have. Session timeouts aren't just a compliance checkbox. These are safety mechanisms.
Cultural Context Determines Adoption
We could build the most technically perfect system in the world and it wouldn't matter if the people using it don't trust it or don't see themselves in it. Working in the Pacific taught us that cultural competence isn't a "nice to have" phase in the design process. It's the foundation.
Language matters. Not just translation, but tone, terminology, the way questions are framed. Visual design matters. Colours, imagery, layout conventions that feel familiar rather than foreign. Workflow matters. The way a case worker in Samoa processes a case is not the same as how it would be done in Wellington, and neither is wrong.
We worked closely with local staff throughout the design process. Not as consultants reviewing wireframes, but as partners shaping the fundamental structure. Their knowledge of how support services actually operate on the ground was irreplaceable.
What Standard UX Gets Wrong
Most UX processes assume a user who is calm, seated, well-lit, and connected to reliable internet. In crisis contexts, none of those assumptions hold.
Connectivity. Internet access is variable. The system needs to work offline or at least degrade gracefully. Losing a partially completed case file because the connection dropped is not acceptable.
Environment. The case worker might be in a shelter, a community centre, or someone's home. They might be on a phone, not a desktop. The interface needs to work in unpredictable conditions.
Time pressure. Sometimes a case needs to be documented quickly because the situation is developing. Long, multi-step workflows don't work when someone is waiting.
Emotional state. Both the case worker and the client may be emotionally affected. Error messages that are confusing or forms that lose data don't just create frustration. They erode trust in a system that people need to rely on.
The Responsibility
This kind of work changes how you think about everything else you build. Once you've designed for a context where bad UX has real consequences for real people, it's hard to go back to treating design as decoration. Every interface is used by a person with needs, constraints, and a context you might not fully understand. The crisis work just makes that truth impossible to ignore.