The first phase focused on understanding how IPM was used, or avoided, in real border operations through Sandbox Experiments. We worked with WCO teams and selected member administrations to clarify user groups and contexts. Frontline inspection officers needed to check shipments quickly, often in environments with unreliable connectivity, mixed device fleets and limited time per inspection. Intelligence analysts required structured access to historical cases, seizure patterns and rights holder alerts. Brand protection and legal teams on the rights holder side needed clear paths to file information and review enforcement activity.

Through interviews, workflow mapping and remote observation, we documented concrete barriers to use. Officers reported that key tasks required several screens and frequent switching between sections. Training teams highlighted the effort needed to onboard new users across multilingual environments. Member administrations described parallel spreadsheets and email chains that had emerged around the platform, a typical symptom of a border management system that is seen as hard to use.

This research allowed us to frame adoption issues in operational terms rather than abstract usability language. It also gave product leaders at WCO evidence to prioritise core inspection workflows and rights holder alert flows before secondary features. The findings created a shared view of the problem among operational units, IT teams and programme leadership, and gave us a stable base for information architecture and workflow restructuring.

Based on these insights, we rebuilt the information architecture around real inspection and case management flows rather than internal structures. The goal was not only better navigation but measurable reduction of mental effort during time critical tasks. For inspection officers this meant that core workflows, such as checking a shipment against alerts and recording an outcome, could be completed in fewer steps with clearer system feedback. For analysts and rights holders, it meant more predictable access to cases, alerts and documentation without extensive searching.

Cognitive principles guided many detailed decisions through tension-driven reasoning. We favoured recognition over recall by making key categories and actions visible at each step. We reduced the number of choices per screen to shorten decision time, especially in multilingual border management interface scenarios. We used progressive disclosure so that advanced options appear when relevant, not as a permanent source of noise. Hidden educational micro features, such as contextual hints that appear the first time a complex action is performed, helped users acquire efficient patterns without formal training.

The restructured information architecture also reflected the diversity of member states. Terminology was normalised while allowing for local variations, and documentation was written so that it could be translated and reused in different administrations. For WCO this provided a clearer model of the product that could be used in future procurement, technical governance and training work, and it aligned closely with the expectations of a security platform interface design effort.