After the turbulence brought on by COVID-19, the airline industry was not in the best shape. Unsurprisingly, negative sentiment grew as customers faced rising ticket prices, insufficient capacity, long queues, and delayed flights. To address this constantly evolving situation, airlines were compelled to review traditional approaches to gathering metrics for evaluating customer feedback—and these insights hold true today. And with advances in AI, airlines have more tools at their disposal to really listen to what their customers have to say.
Conventional metrics like Net Promoter Scores (NPS) and Customer Satisfaction (CSAT) scores that rate customer experiences on a scale of 1 to 10 can only shed so much light. While these metrics are foundational, meaningful customer feedback goes beyond what they can deliver.
The real voice of the customer often lies in narratives—real-life stories that are not found on a form. And personal customer stories tend to shape customer-experience transformations more effectively as they are harder to disregard, particularly when they contain details and specific feedback. By contrast, NPS or CSAT scores are often debated and might not lead to immediate remedial action. Used together, conventional metrics and customer stories can complement each other powerfully.
The following seven approaches may help airlines delve deeper into customer feedback and gain better insight into the state of customer satisfaction. An airline in Asia that implemented these strategies using a Large Language Model (LLM) saw its Customer Satisfaction Index (CSI) go up by more than 15 percent and its contact per passenger drop by more than 30 percent as key pain points were eliminated.
- Maximize existing customer voice data: Detailed insights can be buried in the vast number of customer interactions from contact centers, emails, live chats, and social media. The ratio of contacts per customer to contact centers is typically higher than the response rate on post-experience surveys. Learning how to consolidate and analyze data across these channels is a powerful way to consistently understand customer experience and satisfaction.
- Value customer stories over numbers: Airlines’ post-experience surveys have a response rate of around 5 percent, in line with other industries. Often, they are packed with information. Of those respondents, around 10 percent are ready to share detailed personal stories if given a chance (Exhibit 1). This can amount to thousands of stories that carry valuable insights, especially for major airlines.
- Use advanced analytics to gain precise insights: Reading thousands of customer text messages from call centers and post-flight surveys is harder than looking at a 1 to 10 scale. Advanced analytics tools can help distill insights from text. For example, call centers tend to sort messages into generic categories like “complaints about loyalty program”. LLM-based tools can help to replace those categories with ones that address much more specific customer questions, such as “where are my loyalty points after my last flight?”. Problems at this level of specificity have clear owners in an organization and are easier to action. LLMs can also read customer feedback in various languages, which is especially important for airlines working in multiple markets.
- Prioritize a smooth journey over “wow” moments: Airlines could benefit from getting the basics right first—by minimizing journey friction—before focusing on creating “wow” moments Challenges like disorganized boarding or inadequate communication about flight delays can quickly turn customers into detractors, overshadowing any positive “wow” features like being personally greeted by airline staff.
- Use route-specific strategies: The CSI for highly competitive routes is harder to improve upon than that for less-competitive ones. Differentiating products and services on routes with intense competition from those on routes with less competition can result in an improvement in CSI without the consequences of heavy cost (Exhibit 2). For example, serving premium food on more competitive routes instead of standard fare could be one form of differentiation.
- Develop personalization use-cases: Airlines can use existing customer data to develop personalized marketing promotions that foster stronger customer engagement and encourage repeat purchases. For instance, one airline was able to use flight histories to develop a predictive model that could estimate the likelihood of customers flying from point A to B within the next few months, with about 85 percent accuracy. This was more than enough to help the airline’s marketing team put together a promotional program targeting these customers.
- Inject agents of change at critical touchpoints: Passengers are handed from one organization to the next throughout their journey: from airlines, to ground handlers at check-in desks and boarding gates, to airports. In theory, each company in this chain should strive to deliver the best customer experience, but in reality, the execution machinery often stumbles. Having “agents of change” at critical touchpoints to represent the airline as the end provider of customer service might be useful, costs withstanding of course. Agents could keep an eye on KPIs, such as a CSI for the boarding process or check-in efficiency.
Merging quantitative metrics with qualitative insights may help airlines fully grasp what their customers are experiencing. By focusing on actionable AI-based data intelligence and taking specific actions to listen to their customers’ voices, airlines can surpass customer expectations, build resilience, and secure future growth.
Ryan Jackdy and Stas Melnikov are also authors of this blog post. The authors wish to thank Dicky Salim, Ekaterina Stepanova, Felicia Puspitasari, Garett Hopper, and Vishal Agarwal for their contributions.