Outcomes
Right feature, right people
Because the tourism content came from an API tied to confirmed Singapore bookings, the feature could only activate for travellers who already had a flight booked to Singapore, which shaped both the technical approach and how we measured success.
40%
of travellers flying into Singapore engaged with the feature once it activated on their confirmed booking: a strong adoption rate for a net-new, opt-in discovery feature that nobody had to go looking for.
~32%
of engaged travellers returned to the feature more than once across their trip (pre-flight, day-of, and post-flight), reinforcing that placing it in "My Trips" kept it relevant throughout, not just at the point of booking.
~18%
of engaged travellers tapped through to a listing's phone number, Google Maps, or website: real intent to visit, and the closest signal we had to the Singapore Tourism Board's own goal of driving footfall to local businesses.
Adoption is measured among users with a confirmed booking to Singapore (the only segment the feature could activate for), and is confirmed against deployment data. Return-visit and outlink-click figures are approximate, pending final confirmation.
Future enhancements
Where this could go next
A few enhancements were scoped but not built:
Collaborative planning
Saving and sharing activities and places with travel companions, turning solo browsing into collaborative trip planning.
Local partnerships
Partnerships with local services in other destinations, extending the same listings beyond Singapore.
Personalisation
Personalised recommendations based on users' travel history, e.g. solo vs. family travellers, previous activities, dietary preferences.
Next steps & reflection
Reflections
This project sharpened how I think about designing for a moment that hasn't happened yet. We weren't optimising an existing behaviour; we were trying to rebuild a traveller's confidence and desire to travel, from inside an app they'd open for entirely different reasons, like managing a booking or checking in.
Relying on an external API meant designing for uncertainty from day one, without that ever becoming the user's problem to notice. Gating activation to a confirmed booking felt like a constraint at first, but it turned out to be the right choice, keeping the feature relevant to everyone who saw it.
This work reinforced how much a simple discovery feature depends on getting the surrounding decisions right: where it sits in the journey and what data it can rely on.