Product Scale
Visual Design
Design Systems
Legacy UX
Design Strategy
Role
UX designer; led the design and testing
Scope
Homepage, review page, checkout
Team
Me and my design manager, with partners across dev, product and 4 commercial domains
Timeline
6 months
Outcome
Shipped with the rebrand; review-page conversion held in A/B tests, and travellers found the site calmer, clearer and more trustworthy
Status
Shipped
In short
In late 2025, Webjet launched the biggest rebrand in its 27-year history. I worked on the part the press didn’t cover: bringing the new brand into the pages where people actually book.
There was no design system, the brand guidelines weren’t finished, and the site carried a decade of legacy. Rather than wait, we used the homepage, review page and checkout to test the brand on real screens, and built components as each page needed them. Travellers found the new site calmer and more trustworthy, but price still decided whether they booked, which shaped what I pitched next.
001
The launch in the press. Coverage in news.com.au, Mumbrella and more
The problem
After the 2024 demerger, Webjet wanted to be seen as a travel companion, not just a place to buy flights. Years of iteration had left inconsistent components and pages where everything competed for attention.
The usual order is brand, then design system, then pages. There wasn’t time, so we did all three at once, alongside our regular roadmap. That showed us early how the brand worked in the product, but left us nothing fixed to build on.
002
The new Webjet mark. The rebrand gave us a new logo and identity, but the guidelines for using it in the product weren’t finished yet.
The brand on real components. Navigation, search, the date picker and deal cards, designed with real content to see where the new brand worked in the product and where it needed adjusting.
Designing with an unfinished brand
We defined type, colour, spacing and components from what the three priority pages needed, and adjusted the brand where it didn’t work in the product. There was no clean system on day one. Component by component, we chose what to replace now and what to keep for later, so old and new sat side by side for a while.
003
Deal cards in the new mosaic layout. Patterns like these came out of the priority pages and became the type, colour, spacing and component rules we reused across them.
The homepage
The homepage was loud, mostly from promotional banners, and every team wanted something different. Brand wanted a bold change, engineering wanted minimal disruption, commercial wanted no drop in conversion, and compliance needed regulated pricing shown correctly.
I designed a modular structure instead of one fixed page. Each travel vertical got its own homepage, built from shared, CMS-managed modules with pricing rules built in from the start. I worked through the module architecture with engineering while designing, which let the rebrand go further without adding cost.
For banners, we proposed one prominent slot partners could pay for, instead of equal space for every campaign. Marketing kept room for every campaign and gained a paid placement. Each vertical now runs its own campaigns without waiting on engineering, and marketing builds banners in roughly half the time.
004
The homepage, before and after. Drag to compare the old homepage with the rebranded one.
The review page
The review page is the last step before payment. The old one had everything people needed but no structure, so I rebuilt it around the five questions people ask at that moment: What am I booking? What does it cost? What’s included? What do I need to know? What do I do next?
The harder part was ownership. A hotel card belongs to the Land team, a flight card to Flights, and the payment breakdown to Payments, each on its own roadmap. A change that looked simple in Figma could need three or four teams to ship. So every card had to stand on its own and be buildable by its team, while still reading as one page. It shipped in stages, starting with an A/B test in New Zealand in September.
005
The review page, before and after. Drag to compare. The new page is organised around the questions people ask just before they pay.
What the A/B tests showed
On the review page, conversion held steady against the old page in both markets, and in Australia more people continued to the next step. On checkout, conversion held in New Zealand but dipped slightly in Australia.
Optional extras dropped: insurance, hotel and paid-seat add-ons. That was a trade-off I’d argued for. Webjet’s booking price guarantee holds a fare for 30 minutes, and the old page filled that window with upsells. I’d proposed moving them to after booking. That wasn’t approved, but the new structure still made them easier to skip.
006
What happened after launch
In the first month, we ran AI-moderated interviews with 144 travellers in Australia and New Zealand. They found the site more modern, calmer and easier to read, more approachable for less confident bookers, and more trustworthy, or “less scammy”. The homepage read as a broader travel hub, not just flights.
Behaviour was more mixed. Hotjar showed most people still went straight to flight search and barely scrolled the homepage. On the review page, many removed and re-searched flights to find the true final price. The rebrand changed how people felt about Webjet, but price and fees still decided whether they booked.
That led to my next pitch at quarterly planning: personalised add-ons, using AI and booking behaviour to show each customer only the extras relevant to them. It was adopted as a smaller step, with business rules and A/B tests. The tests didn’t show a clear win, so it went back to discovery, which is where I’d argued it should start.
007
Reflection
We made it work, but it wasn’t efficient. The components we built for these pages never became a maintained design system, and that gap is still there. Next time I’d push for a minimum foundation first: core type and colour, a few components, someone who owns them, and brand basics locked before product work starts.
I’d also agree on measurement before launch. We had A/B tests on the review page and checkout, but none on the homepage, and Hotjar only sampled part of our traffic. Next time I’d agree two or three metrics with the product manager and analyst at the start, so we know exactly what the redesign changed.
008
The rebrand campaign. The launch was led by campaigns like this one. My part was carrying the brand into the pages where people actually book.