June 12, 2018
How Artificial Intelligence Will Make Hotel Distribution Smarter
Because of its vast complexity and reliance on a large quantity of data, hotel distribution is one area of the travel sector that has the potential to be greatly improved by artificial intelligence.
In today’s hotel distribution landscape, inventory information such as rates and availability data are passed from one party to another through a variety of methods, including manual data entry, channel manager, and direct connectivity. As a result, the hotel distribution systems today are like a bunch of old and new machines duct-taped together.
Artificial intelligence has the potential to change this by improving communications, discovery, and customer service among travel providers, travel agents, and travelers they’re looking to reach. Artificial intelligence also creates an incentive to consolidate data. And while the current AI technology is still in an early stage, it’s moving faster than many realize. Soon, thanks to Google Cloud AutoML and other efforts, companies of all sizes will be putting cutting-edge AI algorithms to use.
SkiftX spoke to Wei Xia, vice president of new products and services at DerbySoft, a company that provides high-performance distribution services across the hospitality sector, about how artificial intelligence is already being implemented into hotel distribution and the role it might play in the future.
SkiftX: How could artificial intelligence change hotel distribution?
Xia: Artificial intelligence will enhance visual search, relevant suggestions, and caching to present the customer with the right results at the right time. Established players like Google and Booking.com, as well as startups like Hopper, are already using the technology to innovate and improve distribution.
Artificial intelligence could help improve the customer experience by making the booking process more intelligent. At the most recent Google I/O annual developer conference, Google CEO Sundar Pichai demonstrated Google Duplex, a voice assistant powered by artificial intelligence that is able to make phone calls and have full conversations to carry out simple tasks such as scheduling appointments on the user’s behalf. It’s not unimaginable that a similar service could make travel arrangements on behalf of travelers in future, without them having to click on 10 different websites.
Artificial intelligence can also improve hotel content. For example, hotel content can be translated to a user’s language by using an automatic image recognition algorithm to tag and rate an image.
Here at DerbySoft, artificial intelligence and machine learning has been at the core of our customer segmentation and hotel segmentation model. These technologies are helping us compensate for insufficient data that has plagued the traditional systems to tackle the long tail distribution issue and still offer high performance for the hotels we work with.
SkiftX: How big is artificial intelligence in terms of hotel distribution right now? Is it just starting to impact these systems, or is it already having a major impact?
Xia: Artificial intelligence is already impacting hotel distribution, but it’s still in the beginning phases. Most companies are still learning and experimenting with improving its efficiency, and in order to to improve artificial intelligence when it comes to hotel distribution and content, the AI algorithm must first consume enough tagged data to “get smarter.” A number of innovative startups have already come up with new ways to distribute travel products using artificial intelligence, including our team at DerbySoft. Google is making major inroads as well.
Currently, artificial intelligence is good enough to augment and assist most travelers to perform simple tasks, like checking a flight status on Amazon Alexa, for example. Artificial intelligence will soon power recommendation engines, fraud detection, price prediction, customer review sentiment analysis, and semi-automated customer support systems. Some of these things are already happening today, but there’s still a long way to go.
SkiftX: How do you think artificial intelligence could help optimize a company’s pay-per-click campaign?
Xia: Most consumers still begin researching and booking travel on Google, so pay-per-click marketing is still very important for traffic and leads. Artificial intelligence will power better decision making around pay-per-click campaigns and optimize marketing campaigns by targeting the right audience at the right time on Google and other search engines.
At DerbySoft, we analyze demand and offer (as expressed in terms of placement and user cohorts, and hotel types and their inventory), forecast the essential metrics of the marketing funnel for different combinations of offers, and make bidding decisions by adjusting exposure and volume levels in order to maximize our customers’ objectives.
SkiftX: How will artificial intelligence make metasearch more effective?
Xia: An ideal metasearch campaign is one where the user can find exactly what he or she needs when he or she needs it. Artificial intelligence is already playing a significant role in metasearch ad ranking logic. All of the major players, such as TripAdvisor, Trivago, and Google, use it in order to display the most relevant results for each user in a timely manner. Being able to use AI to predict the campaign metrics would enable better insight, decisions, and ultimately stronger ROI for metasearch campaigns.