Find the right hotel quickly without feeling overwhelmed
Trust that the results shown are relevant to them personally
Get personalised recommendations without having to manually filter through hundreds of options
Drive adoption of the AI Recommendations feature — a key differentiator vs competitors
Build user trust in AI-powered suggestions to encourage repeat bookings
Increase booking conversion rate — fewer drop-offs at the search and results stage
The first iteration surfaced 8 major usability issues — most centred around the filter panel overwhelming users before they'd even seen results. Participants consistently skipped filters entirely.
After restructuring the IA and simplifying the filter hierarchy, task completion improved dramatically. Users found their hotel and reached checkout 35% faster than with the original interface.
Booking.com's scale made it hard to know where to start. The sheer volume of features meant every design decision had downstream consequences on other parts of the flow.
IA is everything. Visual polish means nothing if users can't find what they're looking for. Solving navigation and hierarchy first made every other design decision easier.
I'd involve real users earlier in the process — particularly at the wireframe stage — rather than waiting until a polished prototype was ready to test with.