Booking holidays this Cyber Monday? Find out how Airbnb uses Machine Learning to improve its online sales

Airbnb began life as a niche site providing an affordable alternative to traditional hotels during high profile events where local accommodation was scarce. Since then it has revolutionised the travel and hospitality industry by offering a world of alternative and unique places to stay. Much of its success has been attributed to it’s frequently reported use of data science to improve its range of product offerings and enhance the Airbnb experience for both hosts and guests. Airbnb’s application of machine learning has improved its online service and offered a host of new marketing initiatives to capitalize on in a saturated global hospitality market.

“The application to search ranking is one of the biggest machine learning success stories at Airbnb.” (Applying Deep Learning to Airbnb Search)

Anyone who has used Airbnb will know how refreshing it is to interact with it’s search functionality compared to other sites. The accuracy of recommendations and search results help make the usually arduous task of securing accommodation enjoyable, even on a tight budget. The clever use of beautiful interior shots and helpful suggestions of experiences in the local area make it feel more like you’re browsing a personalised travel brochure tailored to your needs, than frantically searching online for the best place to stay in your desired location.

“The company uses a machine-learned search ranking model to personalize results for guests. The model factors in guests' tendencies to click on certain bookings. For example, Airbnb might look at whether customers favor specific types of décor in places they book. The company feeds more than 100 characteristics into the model, which then uses the data to identify patterns and personalize search rankings.” (Ref: Business Insider)

Personalisation has been a priority for marketers and online retailers for some time now. However consumers are demanding more than just personalised communication, they’re looking for products and services that are tailored to them. A Deloitte study found that 36% of consumers expressed an interest in such products or services, while 48% said they’d be willing to wait longer than usual in order to receive it. Airbnb uses ML to connect millions of guests and hosts using personalized criteria, it “...uses host preferences to personalize search results for guests, promoting hosts likely to accept the accommodation request. For example, the model factors in whether hosts prefer to book guests on back-to-back dates or enjoy having gaps in between bookings.” (Ref: Business Insider). As more people use Airbnb, the richer their data-set becomes allowing them to build models based on guest and host interactions.

Here at RE•WORK we’ve been lucky enough to have one of Airbnb’s data science team speak at one of our Women in AI Dinners. Negin Nejati, Senior Data Scientist at Airbnb, discussed the importance of spending time on defining the machine learning problem in the context of the overall product before working on finding solutions. Negin explained that with the quantity of data now available to companies, combined with the access to open source tools, and the low cost of computation, it can be easy to ‘throw the latest technology at any given problem with little preparation, which can lead to overly complex solutions, suboptimal processes, and waste time.’ Negin's team is making sure customers feel supported throughout their Airbnb journey by providing self-help materials to users to solve their problems as well as by building tools to enable agents to be more effective in helping the customers. Negin explained how better understanding the end to end product helped clarify and simplify the problem. In particular, she highlighted two examples: 1) Clarifying the definition of the "user issue" and recognizing that it encapsulates 3 different concepts: "user issue", "root cause", and “desired resolution" helped clarify the problem. 2) Limiting the length of the text customers can send to explain their issues helped simplify it. You can hear more from Negin and her work at Airbnb in Episode 36 of our Women in AI Podcast: AI at Airbnb: Are we solving the right problem?

Interested in learning more about businesses using machine learning and deep learning to improve their online sales and user experience? Use our Cyber Monday discount code: CYBER25 for 25% off your ticket to one of our summits. Offer valid until Friday 30 November and can be used in conjunction with existing offers, why not beat the queue and get a double discount on any Super Early Bird or Early Bird Passes.

Upcoming events:

Women in AI Dinner, San Francisco, 22 January, Deep Learning Summit, San Francisco, 24 - 25 January, Women in AI Dinner, London, 19 March, Deep Learning in Finance Summit, London, 21 - 22 March, Women in AI Dinner, Boston, 20 May, Deep Learning in Healthcare Summit, Boston, 22 - 23 May, Deep Learning for Robotics Summit, Boston, 22 - 23 May ,Deep Learning Summit, Boston, 22 - 23 May.