UX / UI Project

Almosafer

Almosafer is a leading travel brand in the Middle East, delivering end-to-end travel experiences through scalable digital platforms and a strong omni-channel ecosystem. The platform provides access to over a million hotels worldwide and flights from hundreds of airlines, alongside a broad range of travel services, including holiday packages and car rentals.

Almosafer users were juggling separate apps to plan a single trip — booking a flight in one place, finding a hotel in another, then piecing together an itinerary by hand. The travel companion app's job was to close that gap: bring flight, hotel, and itinerary planning into one flow.

But the deeper problem surfaced once testing began: for a meaningful share of users, the challenge wasn't finding a hotel — it was deciding where to travel in the first place. That single reframe shaped the direction of the entire redesign, starting with how search itself begins.

Through workshops I facilitated with key stakeholders, I aligned business goals with user needs and translated insights into clear UX directions. This process enabled actionable initiatives across design, engineering, and business teams, supporting a shared product vision across the team.

Role

I led the product design and UX process as one of four team members over a 4-month project, creating UI visuals and collaborating with product managers, marketing, and development teams. Together, we researched, scoped, planned, executed, defined fixes, and prioritized tasks.

Problem Statements

Users were juggling multiple apps to plan one trip — flights in one place, hotels in another, itinerary details tracked manually. That fragmentation was the real problem to solve — not the vaguer idea that "booking is hard."

Project Goals

The goal is for users to plan their entire trip and make all their reservations, such as flights and hotels, without needing to switch between multiple applications.

Filtering, add-ons, room selection — the three moments where users quietly gave up.

Research: Understanding Users

Across our research sessions — a mix of moderated interviews and unmoderated remote testing — three friction points came up repeatedly: the filter interaction was hard to trust, add-ons and room selection required too much back-and-forth to compare, and neither showed hotel policies or final price early enough — so users hesitated right when they should have felt ready to book. The volume of information on-screen made this worse: even confirming a simple amenity meant scanning through clutter that had nothing to do with the decision at hand.

Two paths, one shared foundation: finding a trip, and holding it together once you have.

Streamlined flow for swift search and seamless trip planning

The flow splits into two core paths from Home: Search (Discover, Trending, For You) for users who don't yet know where they're going, and My Trips for holding everything together once they do — ongoing trips, past trips, and shared planning. That second path exists because of a specific insight from research: users wanted a way to plan and split costs with the people they were traveling with, beyond simply booking for themselves. The Multiple People → Invite → Start Trip branch is a direct answer to that — turning a solo booking flow into one built for group trips from the start.

Functional style, with great attention to detail, ensuring both practicality and aesthetic appeal in every element.

Distinct colour palette

Almosafer brand colours are supported by distinguishable colours to call out important interactions. Using animation, I emphasized movement and the sense of possibility it creates, drawing the user's attention when necessary and adding small moments of delight where appropriate.

The style guide I built emphasizes consistency across all touchpoints, including digital platforms and marketing materials. By defining clear guidelines for typography, colour palettes, and layout, I created a cohesive visual identity that enhances brand recognition and fosters trust, ensuring every interaction resonates with users.

Search, itinerary, and chatbot — wireframed as one continuous thread, not three separate features.

Wireframes

Wireframes prioritized one thing: making sure a user could move from searching a destination to seeing it inside their itinerary without feeling like they'd switched tools. Early low-fi versions treated the chatbot as a generic support layer; testing showed people expected it to already know what trip they were planning — so I tied it into the same trip context (dates, destination, current booking) instead of starting every conversation from zero.

A booking flow with 2,582 results still needs to feel like one decision, not two thousand.

Crafting perfection from every element

With results in the thousands, the filter sidebar had to do real work: distance, rating, and property type up front, with live result counts next to each option so users could narrow down before committing to a click. That immediate feedback loop — see the count change, then decide — is what kept a 2,582-result search from feeling like an unfiltered wall of options. I compared the redesigned filtering experience against the existing sidebar in a 10-day remote, unmoderated prototype test with 164 participants (82 per variant). The redesigned version reduced average filtering time by 31%, increased filter completion by 18%, and led to more users reaching hotel results without abandoning the task.

Research revealed that the problem wasn't finding hotels — it was deciding where to travel in the first place.

Reframing the entry point

Before introducing the mood selector, I evaluated three alternative entry points to the search flow — all three assumed users already had a destination in mind. Research showed the opposite: many Dubai-based users hadn't yet decided where to travel, which made destination-first search a poor fit for early-stage planning.

Alternatives considered

Traditional search only — destination, dates, guests. Dropped: many users hadn't chosen a destination yet, so destination-first search didn't fit early-stage planning.
Popular destinations cards (Maldives, Georgia, Turkey, Thailand). Dropped: users browsed the cards but struggled to connect them to their own intent, which prolonged exploration instead of shortening it.
Large inspiration feed — an Instagram-style, continuously scrolling discovery feed. Dropped: it was engaging but increased browsing time without helping users narrow down a decision.

I shifted the entry point from destination selection to travel intent. Instead of asking "Where do you want to go?" I first asked "What kind of trip are you looking for?" The mood selector (Snowy, Tropical, Mountain, Capital) let users filter by feeling before seeing any results, cutting an entire round of aimless scrolling for users who were motivated to travel but hadn't yet chosen where.

Three months after launch, usage told the same story the research did.

Impact

Weekly active users per user (WAU) rose 20% in the first 3 months after launch, compared to the previous version.
The redesigned filter sidebar, validated in a 164-participant A/B test, reduced average filtering time by 31% and increased filter completion by 18%.

The hardest part wasn't designing any single screen — it was making five separate booking tasks feel like they belonged to the same trip.

Conclusion

Looking back, the biggest lesson from Almosafer was that fragmentation gets solved less by adding an all-in-one dashboard and more by making sure every screen already knows what trip it belongs to. I also intentionally accepted a heavier first-time home screen — mood selector, trending, and for-you all visible at once — over a leaner single search bar, because reducing decision paralysis for undecided travelers mattered more than minimizing initial screen density. If I had another iteration, I'd want to validate how well the mood categories generalize beyond leisure travel, since the current set (Snowy, Tropical, Mountain, Capital) was tuned for one kind of traveler intent and may need rethinking for business or last-minute trips. That's the question I now ask on any multi-step product: does this screen know what came before it, or is it starting from zero?