ETIHADBEYOND BORDERSMobile App · Service & Discovery · UX Research
Before I drew a single screen, I did the homework. This is how I researched two features for the Etihad app — the Business Class chauffeur and an AI flight search — and how that research shaped every decision I made.
I've kept this simple and in plain language so it's easy to follow. Each part ends by pointing at the design choice it led to — so you can always see why a screen looks the way it does.
I started with one simple idea: a premium airline should turn its two most stressful moments — getting to the airport and deciding where to go — into moments that feel easy and reassuring. So I wrote down the questions I needed to answer first.
I followed the classic “double diamond” way of working: open up wide to explore, then narrow down to a clear answer — twice. First to find the right problem, then to build the right solution.
My route from a blank brief to finished prototypes.
I leaned on a handful of well-known UX and design methods, working in loops — build a little, review, refine. Here's each one in a sentence, and what I used it for.
Start by understanding real people and their needs, then define the problem, explore ideas, prototype, and improve.
Explore wide, then narrow — twice. First to find the right problem, then to build the right solution.
Study how rivals and best-in-class apps solve the same job, to spot the gap I can own.
Design for the real goal behind the task — what the person is actually trying to achieve — not just the screen.
Picture the actual travellers, then walk their journey to find the moments that hurt.
Expert-review my own screens against 10 well-known usability rules, then fix what I find.
Build reusable, consistent components so every screen feels like one Etihad — the Beyond Borders system.
Design for everyone from the first sketch — strong contrast, keyboard use and screen-reader support built in.
Two big changes shaped my thinking. People now expect their airline app to feel as live as Uber — and they increasingly want to just talk to apps instead of filling in forms.
When a car is coming for me, I expect to see it move, know when it'll arrive, message the driver, and share my trip. A free Business Class chauffeur shouldn't feel worse than a normal taxi app; it should feel better.
Thanks to AI chat, people increasingly start with an idea (“somewhere warm in Europe, under budget, for two”) rather than a fixed route. Flight search still demands exact dates and airports, though, so there's a real gap between how people think and how the tools work.
I looked at three kinds of apps people already compare us to: everyday taxi apps (which set the bar for tracking), premium chauffeur services (which set the bar for service), and other airlines (who I compete with head-on). Here's how they stack up — and where I saw my opening.
I mapped rivals on the two things that matter most for this feature. The top-right corner — great tracking and a premium feel — was wide open. That's where I aimed.
| Who | Easy to find the perk? | Easy to book? | Knows your flight? | Live tracking? | Feels reassuring? |
|---|---|---|---|---|---|
| Uber / Careem everyday taxis |
n/a | Very easy | No | The best | OK, not premium |
| Blacklane / Wheely premium chauffeur |
n/a | A bit form-heavy | Yes | Good | Very |
| Other airlines free transfers |
Hard to find | Often clunky | Sometimes | Usually none | Mixed |
| My design | On the trip card | 5 simple steps | Yes, by default | Uber-level | Built-in |
For search I compared two camps: the classic search engines (fast and trustworthy, but form-first) and the new AI chatbots (great at conversation, but often a “black box” you can't trust or actually book from). Again, the best corner was empty.
Classic search is trusted but not conversational. AI chatbots are conversational but not trusted. I aimed for both at once — natural language and transparency.
| Who | Plain-language search? | Shows what it understood? | Good for ideas? | Can you trust it? | Can you book it? |
|---|---|---|---|---|---|
| Google Flights classic search |
Form-first | n/a | Some | Yes | Sends you away |
| Skyscanner / Hopper search + inspire |
A little | Some | Strong | OK | Yes |
| AI chatbots OTA bots & ChatGPT-style |
Excellent | Often hidden | Strong | Risky | Rarely real |
| My design | Type or speak | Shows & lets you edit | “Anywhere” built in | You stay in control | Real fares |
The key thing I learned: people don't distrust AI because it's weak. They distrust it because they can't see what it's doing. So I built my AI around a simple, honest loop:
In plain words, or by voice.
And you can edit any part.
Real flights, with reasons.
I pictured three kinds of travellers and, for each, the real goal behind what they're doing. (That goal, the “job to be done,” is what actually shaped my decisions, more than any single button or label.) These are drawn from my experience and research, not live interviews. I deliberately included an older traveller too, since designing for them (clear text, simple steps, reassurance) makes the app easier for everyone.
“I fly Business so I don't have to think about logistics. And when I'm planning a trip of my own, I just want to say what I want and trust the answer.”
Get from her door to the gate without stress on a work trip — and turn a rough travel wish into a plan she believes in when it's personal.
“I need to know it's handled. Show me the details, let me change anything, and don't surprise me later.”
See and steer whatever the app does for her, so she knows it won't surprise her.
“I've flown for forty years. I value good service and a clear screen — I don't want to fight with an app.”
Travel in comfort and arrive confident, without the technology getting in the way.
I walked through each journey as it works now and tracked how the traveller feels at each step. The red, unhappy moments are exactly where good design earns the most trust — so that's where I focused.
| Step | Find the perk | Book the car | Wait (on the day) | The pickup | To the airport |
|---|---|---|---|---|---|
| What happens | Hunts through menus / website | Fills a form, picks a time blind | No live status — keeps checking | “Is this even my car?” | Watches the clock and traffic |
| How it feels | Unaware | Unsure | Anxious | Wary | Tense |
| My fix | Show it on the trip card | Suggest the right time | Live ETA & status | Driver, photo, plate, contact | “We're watching your flight” |
| Step | Have a rough idea | Turn it into a search | Dig through results | Compare & trust | Book |
|---|---|---|---|---|---|
| What happens | “Somewhere warm, ~$700, for two” | Forces it into date/airport boxes | Sorts, filters, opens 10 tabs | Doubts the AI / the price | Re-types everything at checkout |
| How it feels | Excited | Frustrated | Overwhelmed | Distrustful | Cautious |
| My fix | Accept plain language | Show what I understood | Rank with a clear “why” | Editable details + clear, upfront fares | Straight into checkout |
Everything above boiled down to six simple lessons. Each one became a problem I had to solve in the design.
On the day of travel, people don't want more information. They want to feel looked after. A moving car and a named driver beat any paragraph of policy.
The free chauffeur is generous, but it's buried, so people don't use it. Finding it and booking it should be the same, easy action.
Travellers start with a vibe; search demands exact airports and dates. I needed to accept the vibe and gently turn it into a real search.
The blocker isn't that AI is bad. It's that people can't see what it did. Show the thinking, let them fix it, and trust follows.
Adding AI mustn't get in the way of someone who just wants a normal search. Both doors stay open.
Letting people pick exact times and add extras made the screens busy fast. The fix was to lead with the AI's recommendation and reveal the extra controls only when they're wanted.
I turned those lessons into eight plain rules. Every screen I designed follows at least one of them.
Lead with status and human touches. Calm is the feature.
Use the flight to suggest times, watch for changes, and warn before things get risky.
Show what it understood, let people edit it, explain the “why”. Control stays with the person.
Luxury is smart defaults and no payment fuss, not decoration.
Use familiar patterns (live tracking, chat), then make them feel a notch better.
Design for delays, tight times, full cars and changed minds. Confirm before anything you can't undo.
Show the recommendation; keep the extra controls one tap away. Calm by default, powerful when wanted.
This is the part I'm proudest of: every design choice traces back to something I learned. Nothing here is a guess.
| What I learned | What I designed | Where you'll see it |
|---|---|---|
| The perk is hiddenChauffeur | A “Chauffeur included” row right on the trip card — find it and start in one tap | Home screen |
| Calm beats cleverChauffeur | A live map with countdown, driver photo, plate and one-tap call/message | Live tracking |
| Always know the flightChauffeur | A suggested pickup (“4h before”), the option to set your own time, a clear warning if it's too tight, and “we're watching your flight” | Time · review · confirm |
| The right car for the groupChauffeur | Larger parties auto-upgrade to a BMW X7, and the vehicle card updates in place so it always matches the notice | Passengers |
| Ask at the right momentChauffeur | The “how was your ride?” rating appears only after the trip completes — never before | Arrival |
| No dead endsChauffeur | Online check-in is shown but disabled until it opens; the contact number is editable and validated | Home & passengers |
| People think in ideasAI search | A plain-language box you can type or speak into (real voice input), with example prompts and a first-run hint as the way in | Home & prompt |
| AI must show its workAI search | A “here's what I understood” screen you can edit, a live match count, and a “why this trip” on each result | Understand & results |
| Help the value callAI search | A budget-headroom line (“$88 under your $700 budget”) instead of OTA-style fake discounts | Flight detail |
| Calm over clutterAI search | Curated results, same-day time options tucked behind “change time”, and an honest “fares are indicative” note | Results & detail |
| Extras in the right placeAI search | Optional paid chauffeur, bag or seat offered for Economy only — hidden for Business, which already includes them | Choose fare |
| Don't punish the certain userAI search | A normal “Search flights” option sits right next to the AI one | Home screen |
| Premium = fewer stepsBoth | Smart defaults, no payment for the free chauffeur, real fares straight to checkout | Review · fare · confirm |
| Works for everyoneBoth | Keyboard use, screen-reader labels, gentle motion, strong contrast | Every screen |
My research and experience point the way; real people would confirm it. I haven't run user testing for this concept — so here's exactly how I would, and how I'd know it passed.
Watch people use it — 5–8 travellers per feature, thinking aloud as they go.
A quick “where would you tap?” test — does the new AI button read as “search” without confusing anyone?
Bigger online tests — to measure success rate and speed at scale.
A day-in-the-life check — follow the chauffeur experience on a real travel day.
“You're flying Business on Friday — arrange your free chauffeur to get there in good time.”
“Your flight time shifted — set your own pickup time.” (checks the custom-time control + tight-time warning)
“Find a trip anywhere in Europe, under $700, for two.”
“The AI got your budget wrong — fix it, then add an extra bag to an Economy fare.” (checks control + the optional extras)
To be clear, these are the targets I’d hold the design to in testing — not results I’ve measured, since I haven’t run user testing for this concept yet.
complete it without help
the “excellent” range
quicker than the old form
no accidental bookings or mis-set times
“I felt in control of the AI”
more chauffeurs actually booked