Predictive Automated Negotiation & Dynamic Optimal Rate Adaptation
In recent years, the travel industry has undergone a major transformation. The digital revolution has given rise to large booking intermediaries (known as OTAs, short for Online Travel Agencies, i.e. Booking.com, Expedia, …), which have rapidly achieved a dominant position in the market. While facilitating the booking process for travellers, OTAs also have a significant cost for hoteliers, with commissions ranging from 15 to 30% of the final price.
Hotel professionals have recently begun to combat these high commissions, thanks to a trend known as “going direct”. As the name suggests, the main aim of hotels is to slowly move their potential customers from OTAs to their own booking engines. This is far from an easy task, especially given the huge amount of money OTAs spend on marketing, hotels need to look for other ways to attract customers.
PrivateDeal seeks to help hoteliers achieve this goal and has developed the first intelligent negotiation solution that allows guests to submit their own price for a room. Upon receipt of the offer, the system automatically and instantly negotiates with the guest, if necessary, the best price for both parties.
Although the system has already proven effective with our current hotel partners, there are some challenges that need to be addressed before PrivateDeal can reach its full potential. One of the main difficulties with the current iteration of the product is the need for hoteliers to manually adjust the discounted price via their extranet. To add some context, today, hotels have to manually determine, for each room category over time, the negotiation margin given to the system to negotiate with guests. It is often very difficult and time consuming to find the ideal discounted rate. A combination of many data is needed to optimise the discount, such as the customer’s price request (collected by PrivateDeal), room occupancy, customer booking windows, competition, etc.
It is on this theme that the Icare Institute proposes solutions by collecting the different data needed (External Crawling) and by developing algorithms combining machine learning techniques and heuristic methods. The number of elements that influence the price of a room, the large number of different types of hotels, but also the diversity of customer profiles, their preferences, etc. reveal a large number of parameters. Training machine learning models on such a sparse data set is a real scientific challenge that the ICARE institutes have already partially met. The Icare institute also offers its expertise in software and algorithmic engineering by proposing a process that automates and personalises the negotiation phase between the client and the hotel. In the end, there is only one objective: the optimisation of the conversion rate.