CUSTOMER SNAPSHOT

Norwegian is the fifth largest low-cost airline and one of the fastest growing airlines in the world, with approximately 5,200 employees and 23 million passengers in 2025. Around a thousand people work in customer service across phone, live chat, email and social media, day and night. Many passengers get in touch before, during and after their journey, and a significant share of those enquiries arrive through chat. Vegard Andersen, Principal Advisor Service Delivery at Norwegian, leads the work of keeping that operation responsive as volume climbs.

‍
THE CHALLENGE

Norwegian's customer service team was running into a familiar curve. The volume of enquiries was rising, the channel mix was shifting toward chat, and the operation needed to stay open around the clock without scaling the team in lockstep. The status quo was unsustainable.

The team was clear on what they were optimising for. Vegard Andersen puts it simply.

"We want to be available to our customers around the clock and improve the response time in other channels. The chatbot helps us with both."

Their support team would be massively overwhelmed if enquiries continued to pile up at the rate they were arriving. Norwegian decided they need automation that will carry the most frequent enquiries on its own, in the languages its passengers actually used, while giving agents back capacity for the cases that required them.
‍

THE SOLUTION

Norwegian evaluated several suppliers and selected Kindly. Their goal was to find a platform that handled all their data integration and multilingual needs, while requiring the simplest possible setup and maintenance. 

"We chose Kindly because of its simplicity and ease of use. We liked its multilingual capabilities, and the fact that it allowed us to connect easily with other systems." — Edward Thorstad, Director Customer Service, Norwegian

Norwegian also valued that Kindly’s team managed the project directly, without a third party in the middle. The project group brought together expertise across tourism, customer support and language, including specialists in tone of voice and plain language to make sure the AI Support Agent communicated the way Norwegian's customers needed it to.

The build started from a Kindly-developed template for the airline industry, tailored to Norwegian's purpose. This cut the content creation phase down, and freed time for extensive testing before launch. The first phase focused on the highest-volume enquiries, which gave the AI Support Agent its biggest target from day one. Norwegian and English carried the most volume, so those were where the AI Support Agent went to work first. Integration with Norwegian's other support systems followed, giving the team visibility into enquiries, response times and customer satisfaction in one view.

‍‍
THE RESULTS

Norwegian's AI Support Agent reduced the number of live enquiries handled by their team by 20%. Live chats dropped by 30%. Incoming phone calls dropped by 5%. Today, 1 in 5 enquiries in Norwegian are answered automatically by the AI Support Agent.

The shape of the work has changed alongside the volume. With the most frequent enquiries handled before they reach a person, agents have capacity for the complex cases, including custom outreach to customers affected by specific operational events. Andersen pointed to one early test of that capacity. 

"Shortly after we started using the chatbot, 18 of our Boeing 737 MAX aircrafts were grounded. This, of course, led to a number of questions. We quickly updated the chatbot with information related to questions customers had about this topic."

The flow-on effects show up throughout their customer service operation. Waiting times are now shorter for customers, and work-related stress is lower for the support centre. The AI Support Agent also finds bookings and supports relevant additional sales, asking customers to confirm booking information before guiding them to the right options on Norwegian's website.

For massive enterprises like Norwegian, the measure of automation at this scale is not what it removes from their team of support agents, but what it gives back to them: the time and capacity to handle the cases that actually need an agent’s expertise.