How AI helps Japan Airlines personalize the travel experience

By | 3 minute read | July 16, 2019

Japan airlines plane

For airlines, the sheer volume of flights and travelers can sometimes make it difficult to provide a personalized customer experience. When airports are busy and flights are full, passengers sometimes feel that the airline simply sees them as objects to be transported from point A to point B.

In response, Japan Airlines decided to set a new standard in customer service. We wanted to provide a more personalized travel experience by learning from our customers and taking their personalities into account. For assistance, we turned to IBM Services to help us apply AI to the solution.

Better customer engagement with AI

Our first AI project had two objectives. First, we wanted to challenge ourselves to explore the potential of AI solutions. Second, we wanted to engage with our customers in a way that would truly enhance their flying experience with JAL.

We quickly identified travel to Hawaii as a focus for the project. One of our most strategic destinations, Hawaii is around a seven-hour flight from Tokyo. Many travelers—especially parents traveling with young children—were concerned about the long flight, or had questions about vacationing in the islands. We realized that a virtual assistant powered by AI could save our customers and employees a lot of time.

As a pilot project, we built a virtual assistant application called Makana-chan. Based on IBM Watson Assistant and other services from the IBM Cloud including Db2 Warehouse, Makana-chan’s dialog interface understands natural language questions from customers and responds accordingly.

The name Makana-chan comes from the Hawaiian word for “gifts,” and the AI virtual assistant has certainly been a gift for our customers. After we launched the app in December 2016, we found that travelers were pleasantly surprised by Makana-chan’s personality—its friendly, conversational tone helped put them at ease.

We also found the virtual assistant could satisfactorily answer questions on many topics for parents traveling to Hawaii with infants. For example, it could provide advice about the services available for babies at airports, in-flight options for meals and sleeping arrangements, whether babies need passports, how to find medical assistance and where to buy baby food in Hawaii.

Recommendations personalized for each traveler

What makes Makana-chan truly innovative is its personalized recommendations, which are enabled by AI analysis and understanding of our customers’ personalities. These insights come from using IBM Watson Personality Insights to analyze customers’ Facebook and Twitter posts when they log in. Makana-chan can then assign the traveler one of nine personality types and give advice based on their interests and preferences. This level of communication and understanding truly differentiates our solution.

The latest version of Makana-chan also integrates IBM Watson Visual Recognition, which can analyze customers’ photos, detect their age, gender and other characteristics, and use this data to improve the recommendations.

Improved customer satisfaction

Makana-chan has demonstrated how AI can transform customer engagement. As an example, a customer survey showed that more than 70 percent of customers were satisfied with Makana-chan and over 80 percent would recommend the service to others.

What’s next for Makana-chan? We are actively working to expand the virtual assistant’s accuracy and knowledge, to extend it to businesses and consumers, and to integrate it into our travel booking system. That way, the virtual assistant could book seats for customers when they decide to travel to Hawaii and elsewhere.

We are delighted to lead the way in improving the passenger experience through this pioneering use of AI. We look forward to working with IBM to explore other ways that virtual assistants like Makana-chan can benefit JAL and our customers.

Watch Takayuki Okamoto discuss applying AI to transform the airline passenger experience: