Widespread Use of AI in Travel Industry

Bob Offutt, Senior Technology Analyst, Phocuswright, has outlined the opportunities for the effective use of Artificial Intelligence (AI) in the travel business. We began by asking him about the research and how AI can be implemented across the industry.

One of the startling findings in the research is the significant growth in the use of AI without a lot of fanfare. AI applications will define much of how travel will be shopped, distributed and experienced in the next decade.

How will user interfaces change through using AI?

The future is spoken commands or queries. Already, we have seen how Alexa, OK Google and Siri provide high quality results based on AI technology, even when the query is less than clear. As the mobile and car communications channels grow, interactive speech will grow, as well. With Google Home and Amazon’s Alexa, the primary way to dialogue with their supporting AI systems is with interactive speech.

AI CAN BE EFFECTIVE IN DELIVERING CUSTOMISED CONTENT, TRIP AND LOCAL EVENT/ ACTIVITY PLANNING AND SCHEDULING

What impact will AI have on travel planning?

AI can be effective in delivering customised content, trip and local event/activity planning and scheduling. AI can be used to predict weather for the destination as well as intermediate points. AI could also be used to look at historical route performance and the likelihood of having to spend the night in the airport. Finally, AI is ideal for irregular operations and fleet recovery to reposition aircraft in case of a network event. For leisure travel, the opportunity is for AI to use preferences and historical data to recommend vacation opportunities. This is particularly difficult in leisure, where so much destination data is not tagged (free form). This means if a picture of the beach at Maui is not identified as such, the AI system will need to classify it as part of the search process for a Maui vacation. AI can be used to determine the best time to book, projected load factor or occupancy level, and expected cost, among other things.
For business travel, travel managers can use AI to collect and correlate data to develop “standard” trip costs and to define travel policy. AI can help predict when suppliers will have discounted offerings via dynamic pricing models. AI would be very useful in developing travel budgets.

Will AI aid fraud detection?

As technology evolves, so do the tools to commit fraud (and the tools to combat this fraud). Travel merchants experience the loss of approximately 0.4% of revenue to fraud, with smaller agencies being the most vulnerable. AI could detect fraud through unusual pattern detection. This isn’t the whole picture. A significant number of bookings are rejected as being potentially fraudulent. Some of these are actual ly good bookings (false positives). Agencies reported needing to review approximately 7% of the bookings manually, which created another cost. While there is not an industrywide business case, this area is a perfect target for AI.

From a customer engagement perspective, how will AI manifest itself?

Customer engagement is another super opportunity for AI. Some areas where AI can improve customer support/service include: digital concierge, customer selfservice, personalisation and problem anticipation.

Widespread Use of AI