Tailored Travel: Personalized Recommendation System for Airbnb

Scope of Work: User Research, Prototype Design, Onboarding Design

Read Time: 3 mins

Project Overview

Role: Lead UX Designer - Research, Prototyping, High-Fidelity Design
Timeline: 24 hours (November 26 - 27, 2024)
Team: 2 UX Designers, 4 Data Scientists, 2 Software Engineers
Outcome: Airbnb Hackathon 1st Place Winner

Defining the Problem

Simplifying travel planning for first-time users

We were tasked to create a personalized recommendation engine that suggests more relevant travel packages that align with the user’s preferences and travel history.

Our solution introduced a questionnaire during onboarding to generate tailored recommendations — while also addressing decision fatigue and unfamiliarity with the platform.

Research Insights

Secondary research: Learning about Airbnb

Primary audience is aged 18-45
54% of bookings are mobile-based
Users value clean filters, ease of navigation, and data-driven personalization

Focusing on new users became essential: unfamiliarity with the platform amplifies existing frustrations like decision fatigue and overwhelming choices.

User Interviews: Gathering key insight

I interviewed 3 experienced users about their experience with Airbnb’s current system:

Overwhelming options make planning exhausting
Reviews and trustworthy filters are crucial for establishing trust
Users want tailored, stress-free recommendations

This shows that if seasoned users faced certain barriers, new users were likely to experience them more severely.

Proto-Persona Development

Humanizing insights to craft user-centered solutions

Overcoming Obstacles

Tackling misconceptions and constraints for a cohesive design

A key challenge early in the process was advocating for the importance of user-centered design to team members unfamiliar with its value in crafting impactful solutions. I stepped in to clarify its role, addressing divisions within the group, and guiding us toward a unified, user-focused approach.

Another key challenge arose as we quickly progressed to high-fidelity designs. Early collaboration with engineers revealed that our planned "card-stacking" design wasn’t achievable under the time constraints. We immediately pivoted to a simplified layout that balanced user experience and technical constraints.

These efforts were critical in overcoming obstacles and delivering a polished, cohesive solution that reflected our team’s collective strengths.

User Journey

Guiding users from preferences to personalized recommendations

This flow simplified onboarding, addressing user frustrations while aligning recommendations to their needs.

Design Solution

High-fidelity prototypes prioritizing usability and personalization

Our design prioritized ease of use, personalization, and trust-building. Key features included:

Curated Recommendations: Personalized travel packages based on user input
Pre-populated Filters: Preferences from the quiz are applied automatically to save time
Mobile Optimization: A clean, intuitive interface tailored for new users

Winning Strategy

A focused, collaborative approach with immediate impact

Our solution stood out because it combined deep research, design simplicity, and technical alignment. By seamlessly integrating data-driven personalization into onboarding, we made travel planning intuitive and enjoyable for new users.

Our research insights and design decisions harmonized with the data scientists' findings and engineers' constraints, creating a unified and seamless story across research, design, and development.

This project reinforced the power of focused collaboration under time constraints. By prioritizing user needs and clear design, we created a solution that felt both meaningful and impactful.