How Artificial Intelligence is Changing the Travel Industry
From spotting to highlighting the decision-making process for specific demographics to keeping track of luggage (and doubling as an impromptu charging station), Artificial Intelligence (AI) is proving to be a powerful force for the travel business.
Unpredictable, fragmented data — challenge accepted
Of course, AI is making its mark everywhere, but travel is a fairly unique business. First, it’s full of data. Nobody is counting how much toothpaste you use, but every plane seat is monitored, most trains weigh their loads to approximate usage and many cities collect and send out public transport information – so much so that global city transport app, . All that data is ripe for AI interpretation, to create optimised experiences.
Second, travel is full of unpredictability. Not only things like weather (acts of God can change everything) and geopolitics (political unrest can mean that last year’s figures are no predictor of the future); it’s worse than that. As humans, we are deeply inconsistent. Some people want to return to the same cottage for a holiday every year. Others want nothing more than to go somewhere very different and challenging. Understanding the complexities of our interest in travel is the sort of job AI was made for.
Finally, returning to product optimisation, travel is fragmented. A simple airline flight can involve over 50 different service providers, including infrastructure (airports, traffic control) which may be international and public sector. Shaving pennies from operations can have dramatic effects in the long term, but requires immense vision from business information systems.
AI therefore has much to offer, so here are some examples of how it is sweeping through the travel industry:
Pricing and Forecasting Tools
Placing relevant offers to travellers at the right time and at the optimal price (i.e. the best price for their needs – not ‘leaving money on the table’) is the holy grail for travel operators. By combining huge stores of historical data, events tracking and predictive analytics built from sophisticated machine learning algorithms, companies can now meaningfully predict where and when travellers might want to go, automatically generating custom advertisements and boosting conversions all the way down the sales funnel.
, for instance, uses AI technology to create tailored travel offers for its users. With a customer’s permission, Amadeus will scan their social media activity, building a sophisticated model of interests, likes, their social graph, and information from past travels, purchase behaviors and loyalty programs. They then take this collected data and use machine learning to identify patterns from a wider customer community and match users with relevant offers. After scanning for the best deals and interpreting other digital behaviours, such as how the person booked in the past, the offer is placed in optimal positioning for conversion.
One of the true benefits of AI: dealing effectively with complication (what in the discipline of process management is called ‘exceptions’) where in the past, the power of computing has most been felt in optimising run-of-the-mill, standard interactions.
Making business better for corporate travel
Behind the scenes, everyone in the business travel industry from carriers through agencies and the companies that make bookings for their staff want to see efficiencies – but that shouldn’t come at the expense of usability or flexibility.
Tradeshift has revolutionised the procurement and payment function across many sectors, but in 2016 it bought out a business called HyperTravel, whose travel-focused AI technology is now embedded at the heart of , a mobile-first procurement optimisation system.
For corporate travellers, Go gives the simplicity of AI-driven bookings, wrapped up in the careful corporate credit limits, budget allocations and pre-approved credit designed by a company’s HR department. Users can buy, knowing that the numbers are taken care of; and procurement managers can sleep a little easier.
Virtual Customer Service
AI has also been put to use to help create a smoother experience during travel. provide solutions based on natural language processing (NLP), a subfield of AI applications, providing context for the questions that customers ask, so that relevant suggestions can be made.
In addition to Microsoft’s Cortana, Apple’s Siri and Amazon’s Alexa technologies for general travel inquiries and planning, IBM’s Watson is being used by Hilton Hotels to create its travel buddy, Connie, who helps travellers during their stay. Connie works in collaboration with Wayblazer, another Watson-connected travel advice tool, to provide information and suggestions to enhance the traveller’s experience.
In particular, Hilton rightly sees Connie as a fairly capable assistant – an augmentation of its front-of-house staff – but a supremely capable listener. The more guests interact with Connie, the more it learns, adapts and improves its recommendations. The hotel will also have access to a log of the questions asked and Connie’s answers, which will enable improvements to the long-term guest experiences .
Of course, AI can be fun, too. Royal Caribbean Cruises has incorporated chatbots and taken AI implementation even further with its the AI aboard its “Quantum of the Seas”, which travel website notes includes state-of-the-art technologies like “robotic bartenders, virtual balconies and RFID luggage tags.” Some of these are little more than proofs of concepts; but Royal Caribbean is correctly banking on leisure being fertile ground for innovation in AI and VR experiences.
The Skyscanner chatbot is another example of AI technology in use to improve travel customer service. It scans external sources to allow for better scheduling and redirection when plans change as well as making suggestions and answering questions based on room availability, weather, traffic conditions and transportation options.
It capitalises on one of the true benefits of AI: dealing effectively with complication (what in the discipline of process management is called ‘exceptions’) where in the past, the power of computing has most been felt in optimising run-of-the-mill, standard interactions.