Airbnb: proposing intelligent recommendation using AI

Over-whelmed by the hundreds of options and search categories, providing a personalised suggestion bar based on user preferences and booking history.

🎯Objective

Introducing and Enhancing the Recommendations by AI-Powered Personalisation.

📑 Introduction

To enhance the customer experience of Airbnb users by leveraging AI to provide personalised recommendations for stays and experiences, utilizing user preferences and booking history. This will not only boost the engagements but will improve the customer experience resulting in more no. of bookings.

❓ Problem Statement

Currently Airbnb offers a range of stays in vast categories like beachfront, lakefront, mansions and many more but there is little room for improvement. Users are looking for more personalised and tailored recommendations that matches their preferences, aiming for seamless and enjoyable experiences.

🛟 Limitations

General Limitations

• The current system only allows to browse through the broad categories of stays & experiences and doesn’t always hits the mark leaving the customer over-whelmed with the vast number of choices. When it comes to personalized suggestions, it fails to deliver and don’t fully reflect user’s individual preferences.
• Experiences window provides very unique but limited number of options. Luxe experiences are mostly sold out resulting in sub-optimal user experience. Airbnb doesn’t take advantage of experiences feature as it only provides various city tour options and photography packages. Not providing users with a comprehensive overview of the destination.
• No sorting filter , no rewards or discounts or price drop notifications.

🔐 Key Challenge

The challenge involves analyzing the user travel related searches and bookings details. The competitors have an edge over Airbnb, by knowing all the details about the flight bookings, cab bookings for personalizing more accurate recommendations. Airbnb can make use of consented cookie data to improve on the suggestions and past searches related to travel for accurate recommendations.

Business Goal

Goal	Description
Enhance user experience and boost engagement	Make Airbnb more engaging for the user by offering recommendations that feels personal and relevant, encouraging user to explore more of what it offers.
Boost frequency of Stay	By suggesting options that matches user preferences, we aim to encourage them to take more trips and spend more days often.
Build brand loyalty	With the personalised experience, we aim to increase customer satisfaction and loyalty. Making Airbnb their go to choice for all their stays during travel.

🎯 User Goal

Personalised Experience

Users want recommendations that matches their interest and past bookings, making their experience more enjoyable and relevant.

Convenience

All users appreciate is streamlined booking/ selection process without getting overwhelmed or confused by thousands of options. It makes them search what they want, saving time and efforts.

Discover New Services

Users are open to exploring and try out new features like experiences especially when they feel it can add value to their travel. Personalized recommendations of experiences at the booking destination can make the user enjoy the stay better.

Trust and Privacy

Providing top-rated properties with great reviews will harness their trust in the platform and knowing their data is secure and used responsibly will further make them put their trust in Airbnb.

🤵 User Persona

Persona	Pains	Needs
1. Karl (30 yrs) - Seattle, Loves to explore serene destinations.	As a full-time working professional, I need weekend getaways to blow of the steam and is away from the hustle & bustle of the cities.	I would really like personalized recommendations of stays under 100 mile radius. It will help me rejuvenate over the weekends, allowing me to return to work energized and ready to begin anew.
1. Linda (22 yrs) - California, Loves to party and hang-out with her friends.	As a young adult, I mostly travel with my friends on various occasions. I prefer to stay at places which provides fun activities and modes of entertainment.	I would really love the Personalised recommendations of experiences like group tours and amusing activities. Like stays with pool, cabins for treks and camping options and similar.
1. Tom (45 yrs) - Texas, Loves to travel with his kids and take family trips.	Travelling with kids and big family can be overwhelming if you don't consider convenience and entertainment over anything. I want my kids to enjoy and have fun during vacation trips.	I would really love to have Personalised recommendations for kids-friendly stays, having entertainment options and experiences. let's say beaches, amusement parks and playgrounds in the vicinity of the stays.

🏅 Success Criteria

1. Aim to decrease 20% of decision time after Personalised recommendations.
2. Achieve 15% increase in interaction with recommendations.
3. Target 15% increase in the conversion rate after recommendations.
4. Aim for 10% increase in the bookings of experiences by the users who received personal recommendations.

👨‍💻 User Stories

• As a parent, I want recommendations for homely yet spacious stays where kids get comfortable and have convenience and kids-friendly attractions nearby to make our vacation enjoyable.
• As a young new user, I want recommendations feature to provide with fun stays with friends and exciting experiences option like concerts, bar crawls and various shows happening around for enjoyment.
• As a married person, I want the recommendations feature to suggest romantic vacations and getaways at calm and serene destinations for us to spend quality time together.
• As a frequent user, I want the recommendations feature to provide new places and options that matches my past trips/bookings and searches. It will make me explore and have different experiences.

Requirements

Requirement	Details	Importance	Notes
Stay Recommendations	Personalised recommendations of stays based on user's preferences and search & booking history.	HIGH
Experiences Recommendations	Personalised recommendations of experiences based on user's current and past bookings for delightful upcoming experience.	HIGH
Easy Access	The feature will present recommendations in a clear, easy to use format. Help you quickly navigate your options and experiences.	MEDIUM
Preference Filters	User can set his preferences using a filter for more accurate recommendations.	LOW
Data and Feedback	The application will track user behavior, past selections or searches that didn't convert into bookings and user feedback to continually improve on the recommendation and make it more relevant.	MEDIUM

💻 Technical Requirement

1. Recommendation Engine:

• Machine Learning Algorithms: Advanced algorithms to analyze user data and provide personalised recommendations.
• Natural Language Processing (NLP): To understand and interpret user preferences and feedback.

2. User Preferences Management:

• Profile Settings: Options for users to set and update their travel preferences.
• Customization Options: Users can personalise their recommendation experience by specifying preferences.

3. Data Security and Privacy:

• Encryption: Ensuring user data is encrypted and secure.
• Compliance: Adherence to data protection regulations to maintain user trust.

🖱️ Non - Technical Requirements

• User Education: Provide clear instructions on how to use the new personalised recommendation features, including guides and tutorials for users unfamiliar with the functionality.
• Marketing and Communication: Develop a comprehensive plan to inform users about the new feature through in-app notifications, email campaigns, and social media updates to ensure widespread awareness and engagement.
• Customer Support: Train customer support teams to effectively handle queries and issues related to the new recommendations feature, ensuring prompt and accurate assistance.
• Privacy Policy Update: Update the privacy policy and user agreements to reflect the new AI-driven recommendations and data usage practices, ensuring transparency and compliance with data protection regulations.

⏳ User Flow

🧵 Sample Screen

👥 Team Requirement

Team	Requirement	Description
Data Science/MLE	Recommendation Algorithm	Develop and train machine learning models to generate Personalised stays and experiences recommendation based on user data and preferences.
Data Science/MLE	User Behavior Analytics	Implement analytics to track user behavior and interactions with recommendations to refine and improve the AI models.
Engineering	Integration with User Profiles	Ensure seamless integration of recommendation algorithms with user profiles, including preferences and order history.
Engineering	Travel Preference Implementation	Develop functionality for users to set and manage their preferences that the AI models use for recommendation
Engineering	Notification System	Develop and integrate a system to send personalised notifications and prompts about recommended stays and experiences.
Engineering	Performance Optimization	Ensure the recommendation engine and related features do not degrade app performance, including response times and load handling.
Business Intelligence	Analytics Dashboard	Create dashboards to monitor and report on key metrics related to recommendation effectiveness, user engagement, and service utilization.
Quality Assurance	Testing for Accuracy and Privacy	Conduct rigorous testing to ensure the accuracy of recommendations and compliance with user privacy regulations.
Compliance	Privacy and Data Security Compliance	Implement measures to safeguard user data and ensure compliance with relevant privacy laws and regulations.

🫥 UI Requirements

• Personalized Home Screen: Displaying tailored content and recommendations prominently.
• Filters and Preferences Section: Easily accessible filters for travel needs and preferences.
• Feedback and Notification System: Integrated systems for gathering user feedback and delivering personalized notifications.

🏹 Success Metric

Metric	Reason
Increase in User Engagement	Higher user engagement indicates relevance and utility of the recommendations, boosting Airbnb usage.
Improvement in Order Frequency	More frequent bookings suggest the feature encourages additional purchases and improves user retention.
User Satisfaction and Retention	High satisfaction reflects that the feature meets user needs and supports long-term loyalty.
Reduction in Decision-Making Time	Faster decision-making indicates that the feature helps users choose and complete order efficiently.
Accuracy of Recommendations	High accuracy ensures the recommendations are relevant to user preferences, improving user experience.