How did Airbnb become so BIG so quickly?
Airbnb was founded in August 2008 as an online marketplace helping travelers find suitable accommodation. Today it has 4,500,000 listings in over 65,000 cities in 191 countries and offers the widest variety of unique spaces at any price point anywhere in the world.
Based in San Francisco, California, its founders, Brian Chesky and Joe Gebbia, build a platform that helps its users list, discover, and book accommodations and experiences online – with ease and efficiency both for property owners and travelers.
How did Airbnb become so BIG so quickly?
The challenge: serving both hosts and guests
Airbnb deals with two types of customers: hosts and guests. The biggest challenge it faced was collecting data for both categories, understanding them, and finding the optimum way to connect people who offer accommodation to those looking for it.
In early 2011 the company employed just three data scientists, as a small team was able to deal with its data needs. By the end of the same year, Airbnb had ten international offices, and its requirements for data scientists were increased tremendously.
The increasing volume of data and the fast company's growth created the need for the democratization of the work of the data science team, the investment in quicker and more reliable technologies, and the movement of data exploration from the data science team to other groups in the company. Airbnb invested smartly in its teams' education and managed with this strategy to deal successfully with its most significant task: retrieving information from Big Data.
Airbnb collects Big Data (approximately 1.5 petabytes) regarding people's holiday preferences and habits.
Leveraging data-driven decision making
With more than 1.5 million listings in different cities and over 50 million guests, it is not an easy task to serve both parties' needs in real-time. Still, Airbnb knows that customers feedback is pure gold as it helps tremendously in providing better business intelligence.
Decision-making is a complex process based on several facts and situations. Last year the term data-driven decision making appeared in the market after companies and individuals realized how powerful it is to make decisions based on research and related information.
Airbnb wisely used every bit of provided information to build a very successful business model, which paid off its data collection and processing investment.
Business intelligence in practice
To support management in decision making, data scientists collect relevant data, process it, store it and create dashboards to visualize it. The different methods and processes that study various aspects of a business and help improve performance are desrcibed as "business intelligence."
Airbnb tracks all users' actions on their website, trying to understand better the sequence of events that lead up to a decision, e.g., renting a property or canceling accommodation. The insights gained through business intelligence prove that Airbnb follows the correct strategy of listing up property owners in popular destinations during hectic hours. Similarly, data insights can provide a suggested price for accommodation regarding its location, time of year, type of accommodation, transport links, etc.
It is a big challenge to rate and price accommodation, considering the many different types of accommodation, locations, etc. Airbnb has to deal with real houses released in the market in real-time, and they do not follow the same standards as a hotel would. Hence their rating is challenging; additionally, imagine that what is essential for a city center apartment (transport links, Wi-Fi, etc.) might not be critical to a suburban cottage (silence, view, etc.). Nevertheless, Airbnb has developed an algorithm that helps owners price their accommodation based on certain criteria.
A machine learning platform called Aerosolve
The algorithm mentioned above is hosted in Airbnb's machine learning platform called Aerosolve, which can, among other things, analyze properties' photos and provide pricing tips. The algorithm was built based on Airbnb's insights from its customers.
Airpal for Airbnb employees
Airbnb has also developed Airpal, a system that allows all Airbnb employees to access the company's data and tools to analyze data. Airpal is a user-friendly platform that can be used by non-technical users, too, and was successful in empowering the community and freeing data scientists from ad hoc requests.
Addressing online fraud
Airbnb had to earn its customers' trust as users are skeptical of online financial transactions. The company developed customized machine learning algorithms integrated into Airbnb's platform to detect fraudulent transactions. Besides, guests and hosts can rate and comment on others' profiles. This leads to creating a recommendation system that enhances guests' and hosts' confidence and trust.
Airbnb the Greek way
Blueground was founded in 2013 and today is a global team that wants to contribute to real estate. The company offers furnished apartments for a month, a year, or even longer, and the goal is to help people show up and start living.
Blueground has increased, surpassing 4,000 apartments and 500 team members in twelve cities; New York, Los Angeles, San Francisco, Boston, Chicago, Washington D.C., Seattle, Dubai, Istanbul, Paris, London, and Athens.
The secret of success: a fully integrated data science team
Airbnb's most important quality is its ability to transform and react based on its needs. The management soon realized the power of Big Data and the danger hidden in its increasing volume. In Airbnb's example, we see a data science team fully integrated into the corporation and excellent communication with other groups. This was achieved by investing in educating different departments about the power of Big Data, building user-friendly platforms, and a fantastic internal organization of sources.
.jpg)
