Empowering conversational AI & Beyond at Airbnb
Airbnb has transformed how people find and stay in unique accommodations.
However, users still spend hours researching where to go, what to do, and how to plan experiences around their stays.
Timeline
2 - 3 Weeks
Tools
Figma, Figjam, Notion, Adobe after effects, Lovable
Team
Irfan dgli, Rish
Overview
Users often need to jump between multiple platforms , Airbnb for stays, Google Maps for nearby attractions, Pinterest for inspiration, and other tools like Chatgpt, Gemini for itinerary management and general searchs.
This fragmented journey leads to:
Decision fatigue from endless searching and filtering.
Drop-offs between inspiration and booking.
Low engagement with Airbnb’s “Experiences” and “Trips” sections.
Despite having rich user data and travel insights, Airbnb currently lacks a personalized planning layer that understands a user’s intent, preferences, and constraints (like time, budget, or interests) and challenges today’s AI search boom.
Goal
Identify Airbnb’s target audience and identify the problems that can be addressed through the implementation of new conversational AI integration. Also consider ways to connect different use cases and features such as Experiences and services through the feature so users dosent have to juggle between different platforms
Discover
Airbnb
The goal was to understand how users currently plan trips on Airbnb, what challenges they face during this process, and where AI could meaningfully assist without disrupting the existing experience. First, I familiarized myself with the Airbnb app’s features, flows and visual design.
Research methods
I looked at Airbnb’s web and native apps, their community and problem space to briefly map out the research methods to carry out which were.
User Interviews – 10 participants who frequent travelers who use Airbnb 3+ times a year.
Surveys – 120 responses collected via google surveys & Airbnb community forums.
Usability analysis – Mapped the end-to-end Airbnb trip planning journey to identify friction points.
Competitive Bench marking – Reviewed Google Travel, Hopper, Expedia, and Booking.com for AI or personalization features.
Data Insights – Reviewed available data and metrics.
Target audience
Based on their type and content the users can be split into
Primary users - independent travelers - users who value personalized experiences, time efficiency, and seamless technology. They are the majority of Airbnb’s core customer base.
Secondary users - Group & Family Planners - People planning group trips (friends, families, couples)
Other users include remote workers and business trips
User interviews
I conducted user interviews centered on three key questions
How do users user the Airbnb and do they switch between platforms and use or spent much time in the initial process
what are the pinpoints and areas of friction in the process
how do users feel about existing and proposed system
Market research
I did a market research keeping in context the idea of travel and experience companies are undergoing a major shift toward AI-driven personalization. Travelers increasingly expect platforms to act as intelligent assistants, not just booking tools
71% of travelers expect tailored suggestions based on their past behavior and interests (Google Travel Trends, 2025)
Tools like ChatGPT, TripAdvisor’s “AI Trip Builder,” and Expedia’s “Conversational Agent” show rising interest in AI-assisted travel planning
One of the major drive is the users behavioural and habitual approach to search and action has changed due to the AI’s conversational search and integration tools.
Company
Input Controls
Nav Components
Info Components
Containers
Good button hierarchy, clear inputs and useful dropdowns
Great use of tabs, clear side nav bar, search field and excellent icons
Direct alert notifications for certain approvals and helpful popups
Excellent drag and drop cards, well organised modals, clear containers
Inputs missing some refinement. Toggles are clear but not aesthetic
Side nav with search bar, clearly labeled tabs and navigation
Nice tooltips add extra context to the data visualisations
Clearly structured containers and nice cards on the dashboard
Buttons lacking hierarchy and finesse. Input controls need refinement
Lacking clear nav components, some tabs but not much refinement
Lacking sufficient info components, could have more information
Very simple containers with clear structure but lacking complexity
Subtle buttons with nice icons and excellent indicators
Subtle icon side nav bar and clear tabs. Nice search fields
Great progress bars with clear use of colour. Clear notifications
Useful accordions and modals to hide and expand information
Buttons are overly similar, interface lacks visual hierarchy
Multiple tab bars stacked at the top, nav a little confusing
Lacking in sufficient info components, needs more alerts, tooltips etc.
Great for freelancers and small-sized businesses
Define
Persona
I conducted user interviews centered on three key questions
How do users user the Airbnb and do they switch between platforms and use or spent much time in the initial process
what are the pinpoints and areas of friction in the process
how do users feel about existing and proposed system
Mia Geller
Age: 32
Location: Berlin, Germany
Occupation: Software engineer (remote)
Travel Frequency: 9–12 trips per year
Trip Style: Short weekend getaways, city breaks, and remote work stays
Devices: Mobile-first, uses iPhone & MacBook
💭 Goals
Find stays and experiences that fit her mood, budget, and vibe.
Plan and book quickly, without multiple tabs.
Be inspired, not overwhelmed.
😣 Frustrations
“It takes me longer to plan than to actually travel.”
“I wish I could just describe my ideal trip, and it would plan it for me.”
🧩 Motivation Quote
“I don’t want another chatbot. I want a travel companion that just gets me.”
David & Priya Shah
Ages: 39 & 37
Location: London, UK
Occupation: Software Engineer & Marketing Consultant
Travel Frequency: 2–3 family trips per year
Trip Style: Family vacations with kids (ages 6 & 9)
Devices: Desktop for planning, mobile for travel updates
💭 Goals
Simplify family trip planning in one place.
Get personalized suggestions suitable for kids and adults.
Manage multiple preferences (budget, safety, accessibility).
Share trip plans easily with family or friends.
😣 Frustrations
“It’s hard to balance everyone’s opinions.”
“We jump between Airbnb, Google Maps, and WhatsApp to plan.”
“Recommendations don’t feel family-aware.”
🧩 Motivation Quote
“If Airbnb could plan like a family assistant, we’d use it for every trip.”
Problem statement
“How might we design an AI-powered experience within Airbnb that simplifies travel planning, understands user intent, and creates personalized, end to end itineraries, all while maintaining trust, transparency, and user control?”
Affinity mapping
Research inputs
5 user interviews (frequent travelers aged 25–45)
72 survey responses from Airbnb users and travel community forums
Competitor experience audit
Customer journey mapping of existing Airbnb planning process
User journey map
Develop
Low - fidelity
To design this feature inside Airbnb that reduces decision fatigue, adds personalization, and creates a seamless transition from discovery → booking → itinerary.
Design system
Icons - Typography
High fidelity
Retrospective
Visual hierarchy is not cosmetic, it’s functional. Early testing revealed how subtle shifts in contrast or layout dramatically impacted comprehension.
Real user quotes change design decisions. Hearing direct frustrations reframed priorities more than any persona or journey map could.
Balance simplicity with system requirements. While users wanted less clutter, the business demanded visibility into compliance and approvals. Good design meant respecting both.
Design systems are living tools. What started as minimal grew rapidly with barriers to consistency and challenges to achieve order. This was one of the largest learning curves of the project.
Sources and further read
https://medium.com/airbnb-engineering/using-chatbots-to-provide-faster-covid-19-community-support-567c97c5c1c9
https://www.constellationr.com/blog-news/insights/airbnb-look-its-ai-strategy
https://airbnb.tech/ai-ml/intelligent-automation-platform-empowering-conversational-ai-and-beyond-at-airbnb/
https://help.openai.com/en/articles/11487775-connectors-in-chatgpt
https://openai.com/index/introducing-apps-in-chatgpt/
http://newsroom.spotify.com/2025-10-06/spotify-personalized-prompts-chatgpt/
https://www.figma.com/community/plugin/1213050091855586023/figma-gpt