Sam arrived with a complete doctoral thesis, supporting articles and presentations on incident reporting in complex organisations. The first phase focused on translating this material into a product ready conceptual model through domain learning. We read through her main chapters, extracted the key entities and relationships, and mapped how incidents move from initial capture through assessment, escalation and follow up.

In several working sessions we checked our synthesis with her. We grouped recurring patterns from the research, such as incident categories, severity scales, near miss events and follow up actions. These became the backbone of the product language. Instead of a generic reporting tool, we now had a clear model of what must be captured, in which order and with what dependencies.

This phase functioned as evidence based UX design applied to an academic corpus. It also gave Sam a neutral space to separate ideas that belonged in the first version from directions that could wait. The resulting foundations supported prototype design for founders who need to demonstrate both subject matter depth and a realistic plan for execution.

With the conceptual model in place we turned to structure during Sandbox Experiments. The goal was to design information architecture that could handle different incident types without overwhelming users who might only file a report occasionally. In the first week we explored three architecture concepts through option space mapping, ranging from a very linear flow to a more modular approach where reusable blocks appear or disappear based on context.

We used simple flow diagrams to compare the options. Together with Sam we examined how each structure would behave for minor incidents, serious injuries and near miss reports. One example journey started with a short description, then branched into severity selection, body area, immediate actions taken, evidence attachments and supervisor notification. If severity crossed a threshold, additional fields for medical treatment and follow up planning appeared.

After reviewing the trade offs through tension-driven reasoning we chose a modular architecture that keeps a consistent spine but adds conditional sections. This became the UX architecture for mobile apps that need to combine clarity with flexibility. It gave Sam a concrete way to discuss scope with future developers, because it expressed both the stable elements and the configurable parts of the system.

In parallel with the structural work we held dedicated conversations about direction. Sam was a first time founder and wanted to understand how a research based idea becomes a product that can be built, sold and maintained. These sessions were structured, not inspirational. We walked through product strategy for first time founders in her position, using her idea as the reference point.

We discussed several staging options. One focused on a single core flow for incident capture and review inside a small number of pilot organisations. Another added analytics and reporting earlier, which would require more investment but might appeal to certain buyers. We also touched on options for partnerships with safety consultancies and training providers who already work inside companies.

Throughout, we linked each strategic path back to the prototype and to UX design for start ups that operate under resource constraints. By the end of this phase Sam had a simple roadmap of what to build first, what to postpone and how each choice would affect cost, timeline and risk. The design work and the strategic thinking reinforced each other rather than running on separate tracks.

To check that our assumptions held outside the project team, we ran a round of user testing for early products with a small group of people who regularly deal with workplace incidents. Each participant was asked to record a recent or plausible incident, then to review an existing one and add a follow up note. We observed their paths, questions and points of hesitation.

The testing surfaced a few consistent themes. Some users hesitated when the system changed from short to longer forms, so we adjusted the introduction and reordered certain sections to create a clearer sense of progression. Grouping evidence attachments and contextual details together reduced backtracking. Simplifying a few branching rules made it easier for users to understand why new fields appeared.

These refinements did not alter the underlying research but helped align it with observed behavior. For Sam it was a concrete demonstration of how research driven UX and evidence from field users support each other. The updated prototype gave her more confidence when explaining the concept, because she knew it had been exercised by people outside her own circle.