Turning flight search into a trusted decision.

My role

I led UX on Flight Facts end to end, from framing the discovery through to launch and post-launch optimisation.

I planned and ran the user research, designed the site experience, and owned testing and optimisation — in the lab through prototyping and eye-tracking, then live through A/B and MVT experimentation. It was very much a team effort across product, brand and development, but as UX lead I was the decisive voice in the stakeholder group that unpicked which concept — among several competing ideas — would actually differentiate us from competitors: Flight Facts.

Customers were searching flights 39,000 times a day on Netflights.com, but a steady stream of them were picking up the phone anyway. Our answer rate had slipped from 92% to 89.6%, and a quarter of all inbound calls had nothing to do with booking — they were customers asking questions the website itself couldn’t answer: which flight had the extra legroom, which one served a proper meal, which one actually let you watch something on the way.

Servicing those calls cost roughly £15k a year, but the real cost was bigger: every customer who rang to ask a question was a customer we might lose to a rival with a shorter hold queue.

Customers weren’t confused about where to fly. They were unsure which flight to trust.

Our vision, as presented to stakeholders at the time.

Evolve customer understanding

Five renewed personas across the customer base, each anchored to a real attitude towards booking a holiday.

Customer learning — the funnel

We refreshed our customer personas and mapped the full journey funnel — from discovery (“I want to go on holiday”), through information (“I want to go to New York”), to purchase (“I want to buy my holiday to New York”). The gap we kept landing on sat in the information stage: customers had picked a destination, but not yet a flight, and price alone wasn’t enough to decide with confidence.

Building our own database of amenity data across every airline, aircraft and route was considered and ruled out — the scale was enormous, and accuracy would degrade the moment it went stale. Routehappy already maintained that data and scored each flight out of ten, factoring in comfort, speed and quality alongside price. We integrated their scoring with our own systems and branded the result Flight Facts.

Three working prototypes (A, B and C) were built and usability-tested with card sorting and eye-tracking. Card sorting told us the priority order of what customers actually wanted to see first — for tall passengers, legroom outranked everything except price. Eye-tracking confirmed the score was noticed without distracting from the core buying decision: price, airline, and flight time still came first.

The prototypes tested 50% better than the existing Netflights.com site. The clearest late finding was visual, not structural: a traffic-light colour system on the score made the right choice obvious at a glance, and reinforced to customers that they’d picked well.

Routehappy’s own research made the case for why this mattered: at the time, only 38% of US domestic flights had Wi-Fi (a further 9% mid-rollout), and just 13% of regular economy seats offered genuine extra room, while premium legroom options had grown to 40% of domestic flights. Amenity data varied hugely between flights that looked identical on price alone — exactly the decision-critical detail Flight Facts surfaced at the point of search.

The loop that got it there.

Twelve stages, run as one continuous agile process from first sketch to post-launch optimisation.

Showing the fully expanded Flight Facts panel too early in the funnel was a distraction, so we tested a minimised version at the point customers were narrowing down a specific flight, then tested hiding individual elements to see which ones actually earned their place. The fully expanded view, kept for the moment a customer selected a specific flight, reinforced that they’d made the right choice.

Colour as the final lever.

Routehappy’s own scores were branded orange by default, regardless of how a flight scored — a strong number could still look like a warning. We moved to a traffic-light system instead: green for a high score, amber for the middle, red for the low, so the colour matched what it meant at a glance.

The traffic-light colour treatment on the score was tested separately and increased conversion again, worth roughly £0.5m a year on its own — on top of the initial £1.2m from showing Routehappy content, and a further £0.7m from optimising the information architecture. Non-sales enquiry calls fell from 25% to 15% of total call volume, and the answer rate recovered to a consistent 94%.

We believed in the idea enough, backed by the research, to build a full multi-channel brand campaign around it — social, email, paid, display, and a first-ever TV advert: “N is for Netflights.” The campaign helped deliver a 36% rise in mobile and tablet visits, a 30% rise in mobile and tablet bookings, and a 60% jump in click-through on retargeting email.