Swipe-to-Discover Hotel Search Experience
Problem Statement
New users see the same ranked hotels, no matter their taste. Personalization only kicks in from visit two.
top 5% of hotels drive 80% of bookings yet users spend minutes scrolling through irrelevant listings. This leads to fatigue, indecision, and drop-offs.
Weak visual hierarchy for decision-making
Key details like price breakdown, inclusions, and cancellation policy are buried or in small font.
No feedback loop for property relevance
The listing page doesn’t actively capture user sentiment about why a property was skipped or shortlisted.
Filters feel disconnected from results
Filter changes don’t feel impactful as there’s no strong visual cue that the listing set is now“better matched.”
Overloaded with repetitive information
The same property cards keep showing redundant info without progressive disclosure.
Competitor Study & Insights
Airbnb
Great at visual storytelling but lacks semantic, intent-based filters.
Booking.com
Robust filtering but cluttered UI makes personalization harder to spot.
Expedia
Strong bundling and deals, but limited in contextual personalization beyond standard preferences.
Tinder
Masterclass in swipe-based discovery and instant decision-making, but not applied in property search space
Statistics & Data
Our data showed:
Viewing 5 properties → 40% orders Viewing 10 properties → 50% orders
Viewing 20 properties → 70% orders Viewing 30 properties → 80% orders
AI MEETS MAKEMYTRIP
One thing became clear: the more users search for properties, the stronger their booking intent. To harness this, GenAI will be introduced to curate relevant content and reduce the number of properties users need to view before making a decision, using both feedback and intelligent recommendations.”
Expected Impact Scores
⏱️ 40% reduction in time-to-discovery
(Users find relevant stays faster vs. scrolling through 50+ listings in traditional apps)
❤️ 65% higher engagement with property cards
(Swipe interaction is inherently more playful and keeps users browsing longer)
🎯 30% more accurate matches
(AI refinement & semantic search reduce irrelevant results shown to users)
📈 20% projected uplift in booking intent
(Measured by higher right-swipes and“collections”in prototypes)
🔁 25% fewer drop-offs during search
(Because semantic queries and swiping reduce the “search fatigue” factor)
Goal
Reduce time-to-decision for new users by building personalisation into the very first search session.
⏱️
Faster Decisions
+
🎯
Smarter Results
+
💡
Lesser Duration
=
Swipr.AI
The Idea
SwiprAI blends Tinder-style swiping with live AI-powered personalization.
In under 10 swipes, the system learns And updates the hotel list every 3 swipes — instantly reflecting your evolving taste.
What you like (right swipes)
What you avoid (left swipes)
The Experience in Numbers
6th
Smart filter card adapt to what you swipe → Helps refine results in real time
Every 3
swipes Re-ranking initiates → keeps feed relevant in-session
20
hotels are pre-fetched → zero load lag while smooth swiping
How It Works
Semantic Search
AI builds a natural language search prompt from your inputs, learning preferences with every swipe for hyper-relevant hotel results.
Smart Filters
AI instantly adapts filters to match the exact meaning of your typed search queries.
AI Recommendation
The swipe engine instantly adapts to your likes and dislikes, re-ranking hotels on the fly for smarter, more relevant suggestions.
Edit Prompt
Prompt bar stays editable so you can instantly tweak location, budget, or preferences without restarting your search.
Seamless Flow
Effortlessly transition from listing to details to complete bookings faster without breaking the experience.
Collections
All right-swiped hotels are auto-saved in a wishlist stack, ready for easy review anytime.
Search Recall
Quickly resume sessions by instantly viewing and reusing previous searches without starting over.
Special Themes
Occasion-specific themes like Valentine’s personalize results and improve AI accuracy for better property matches.
The Result
The prototype was showcased to the Chief Technology Officer, Chief Marketing Officer, and Head of Design at MakeMyTrip, receiving strong appreciation. It went on to secure the Runner-up position among 20 teams and over 100 participants across the Gurgaon and Bangalore offices.”
The Final Prototype
The User Interface
ONBOARDING
Guides users through a quick setup to personalize their experience.
SEMANTIC SEARCH
Search and refine results semantically through natural voice/text queries.
GenAI STACK
Confirms a new AI-powered stack has been successfully generated.
PROMPT SEARCH REFINING
Helps users fine-tune their prompts & add filters for more accurate results.
COLLECTION STACK
Users view their liked stays & explore similar property stacks.
