Spec before pixels
Information architecture and content come from the Survey Results and About the Island documents (origins, lodging mix, attractions, transport, FAQ). No decorative inventing of services the island does not offer.
This prototype was built to portfolio/production bar—not a decorative mock. Each standard below maps to a concrete UX decision for Taniti tourists and the one-month government timeline.
Information architecture and content come from the Survey Results and About the Island documents (origins, lodging mix, attractions, transport, FAQ). No decorative inventing of services the island does not offer.
Wireframe → guerrilla feedback → interactive prototype → formal peer usability (Task 2). Actionable findings ship; taste-only or out-of-scope requests are deferred with stated reasons.
Book Visit is always available. Lodging and attractions deep-link into the inquiry form so tourists never dump into a single jumbled page without a next step.
Skip link, semantic landmarks, visible focus, aria-expanded menu, aria-current page, and form labels with live success messaging—aligned with WCAG 2.2 perceivable/operable patterns for a public tourism site.
Responsive layout and a real mobile menu address the legacy site’s mobile incompatibility—critical for first-time planners reviewing options on a commute.
Filters, booking deep-links, and form confirmation actually work in the deployed prototype. Placeholders are labeled as inquiries, not invented live inventory backends.
The five Task 2 usability tasks exercise navigation, filtering, deep-link booking, form success, and fact-finding—so peer reviews stress real task success, not visual preference alone.