Working within the Jetmax division of Flight Centre Travel Group, I led a project to improve the conversion rate of customers and redesign the travel booking and management products for Flight Centre leisure brands Aunt Betty and BYOjet.
I kicked things off by conducting a simple heuristic evaluation of the existing booking experience, benchmarked against Jakob Nielsen's ten usability heuristics and presenting my findings back to the product stakeholders. I used the traffic light method of assessing the experience, giving experiences a green, orange, and red rating based on my perception of best practices or UX issues.
The next step was to fully understand who our users were and how they behaved when interacting with our brands. As the division had collected a massive volume of user data over the past few years, there was plenty of ethnographic information to help create a snapshot of who our users were. We quickly discovered Aunt Betty and BYOjet share almost the same user base as they target the same travel markets and rely heavily on traffic from flight metasearch engines like Skyscanner and Kayak.
When I was first employed by the division, the exit rate for users on the flight booking engine ranged from 55-65%. Although many online travel booking websites have a high exit rate due to the fickle nature of customers window shopping, this was clearly higher than usual as almost two-thirds of potential customers weren't converting to customers. We decided to get to the source directly and ask users why they were exiting.
The team collected insights by setting up visitor recordings, heatmaps, and strategic polls throughout the user journey using Hotjar. If a user abandoned their booking on the flight search or checkout steps, a single question poll asked 'Is there anything preventing you from booking?'. We collected thousands of responses with abandonment polls.
My research indicated there was a plethora of reasons why users were abandoning the website. Users were very concerned about baggage allowance, seating availability, the currency type being displayed, additional fees applied to the booking, and additional fees if they needed to change or cancel their flights at a later date. Users also wondered if they could trust our website and whether we were a legitimate operation.
When a user completed a booking, they were prompted to rate their experience with a net promoter score question, followed by several attitudinal questions focussing on why they were travelling and which products interested them. The division collected over 16,000 responses to this poll across all brands, at the time of writing this case study.
I discovered most users were travelling for a holiday or to visit family, some were travelling for personal reasons or for an adventure, and a minority were travelling for business or because we had a good deal. Asking users for insights directly helped set the direction to refine the product offerings and user experience.
When I discovered patterns in the research collected, I attempted to address this by setting up strategic A/B test experiments using the design optimisation tool Optimizely.
One split test experiment conducted clearly showed the country currency type (eg. AUD) on all price totals, which reported a 29.42% increase in conversions for the variation (n=112,450).
Another split test experiment challenged how we ordered travel add-ons in the booking experience, as products offered towards the end of the booking experience had consistently sold less than products offered earlier. I suspected decision fatigue was impacting which products our users added to their cart. I conducted two separate, identical experiments to validate this hypothesis and both returned significant results for the test variation. The first experiment reported a 36.59% increase in users selecting the first travel add-on (n=24,777) and the second identical experiment reported a 50.31% increase in users selecting the first travel add-on (n=30,288).
Although the collective results of the experiments varied from positive to negative, it gave the stakeholders a clearer idea of what aspects of the product needed attention and how we can increase user retention. By listening to user needs, we were also able to reduce the exit rate by 10-20% over the next year and increase the division's reportable revenue by millions of dollars.
When I first joined the team, the only ancillary product Flight Centre Leisure brands offered was extra baggage on flights without baggage allowance. I led the design of new ancillary offerings in the customer experience.
I tested several iterations of ancillary offerings during the flight booking experience and post-booking, with varying success. Ancillaries like paid seat selection and priority customer support converted well during the booking experience, but customers we're more likely to purchase ancillaries like travel insurance and priority support after they'd booked. Customer surveys helped conclude that offering ancillaries at different stages of the user journey increased the likelihood of conversions.
Before coming onboard, there was no one currently responsible for the design of the booking products or the brands themselves – and it sadly showed. There were no rules, principles, guidelines, or consistent UI elements to direct how the products should function, look, and feel. I took it upon myself to work closely with stakeholders, the product team, and the marketing team to own the design space and set about a clearer direction.
To compliment the redesign project, I worked with our division's marketing manager to direct, produce, and animate a series of brand bumpers and destination campaign videos for our more dominant brand Aunt Betty's paid social channels. The fun, colourful campaign helped increase brand engagement and brand equity in the Australian travel space.
Applying the established style guidelines, I then progressively redesigned each aspect of the user interface from flight search to checkout informed by our user research. The design system relied strongly on principles and elements from Google's Material Design system. I designed these interfaces using the UI tool Sketch synced with the Marvel prototyping app for both Aunt Betty and BYOjet brands.
As development resources became scarce, I got my hands dirty and helped contribute to the front-end codebase built on HTML, Sass, and Vue.js for both brands. Below is a sample of the user interface redesign work for the booking and customer account products.