Content area
Full Text
1. Introduction
Airbnb has emerged as one of the stars of the sharing economy. In 2016, Airbnb was estimated to be valued at $30bn (Newcomer, 2016), and in markets like the New York City (Fox, 2016) and Toronto (Gibbs, 2016), Airbnb has been reported to represent 5 per cent of the tourism accommodation revenue. Despite such business success, Airbnb has struggled to optimize pricing by its hosts, potentially missing out on large financial gains. Pricing has been identified as one of the Airbnb’s biggest challenges, and it has been estimated that Airbnb hosts forfeit up to 46 per cent of additional revenues because of inefficient pricing (LearnAirbnb.com, 2015). According to focus groups conducted by Airbnb, hosts become confused when trying to set their prices (Hill, 2015). Unlike the hotel business, which has trained professionals, industry benchmarking reports and technical tools to help set pricing for rooms, Airbnb units are generally managed by regular people with very limited support of pricing tools.
Although Airbnb has been building pricing tools for hosts since 2012, these tools have been very basic and only focused on simple factors related to characteristics of the property, such as number of rooms, neighboring properties and amenities like parking (Hill, 2015). After years of working on a pricing algorithm, Airbnb recently selectively released a new pricing tool that takes both property characteristics and demand into account (Hill, 2015). The pricing of Airbnb listings is unusually complex because, in addition to traditional demand factors such as seasonal changes, local events and location, each Airbnb listing exhibits unique characteristics and hosts often adopt extra roles, such as concierge, cook and tour guide (Hill, 2015). Airbnb’s new pricing algorithm tool therefore uses machine learning to provide hosts with a new pricing suggestion for each date in the future that the host makes the listing available. Once the hosts receive a pricing tip, they can either choose to go higher, lower or do nothing. Thus, unlike many other sharing economy platforms, like Lyft and Uber, where an algorithm controls prices, Airbnb allows its individual hosts to decide whether they want to act on the tool’s advice or not. This individual host control over pricing decisions creates a situation whereby the hosts’ characteristics and skills are as important...