Today’s consumers want a forum to voice their frustrations with air travel because they tend to feel unheard in the heat of the moment. Emerging technology and crowdsourced data could provide them a means of reporting trip outcomes that would allow other travelers to make smarter decisions about which airline to choose before they book flights. The data could also provide airlines with insights they don’t already have. In short, crowdsourced data, analyzed and verified by AI could positively impact the future of air travel.
Why ratings and reviews aren’t all that helpful
Rating systems use rudimentary data: typically, one to four or five stars, which range from awful to excellent. Although the number of stars is intended to convey the degree to which a consumer liked or disliked an experience, their limited range misses important nuances in consumer sentiment.
The challenge with reviews is twofold: First, the information reviewers post can be opposite of a rating when the consumer doesn’t understand that one star means unsatisfactory and five stars means excellent, which is why some one-star reviews read like a testimonial. Conversely, a high rating can be accompanied by a long list of complaints. Second, reviews content is unstructured data that is qualitative, not quantitative, meaning that it can’t be translated into numerical form. As a result, consumers find themselves reading 10, 15, or 20 reviews just to decide whether they want to choose one airline over the others.
Nonnumeric ratings, such as ratings, the number of stars, thumbs up reviews are severely limited in terms of what they can communicate. For instance, they do not allow for consumers to truly compare or contrast the performance ratings of an individual airline against another. Understanding the degree of how one airline is better or worse than another is impossible. However, these are the expectations Gen Z and Millennials have about internet 2.0: easier ways to search, shop and make sound decisions faster.
How aggregated, firsthand consumer data could help
Capturing consumer sentiments is essential if airlines are going to be held accountable for lost baggage, delayed flights and cancellations. That data could also help airlines better optimize both internal and customer-facing services impacting customer experiences. Because the data would be crowdsourced from actual consumers who have recently experienced bad outcomes, the data would be both fresh and reliable.
With each new entry, AI could adjust rankings dynamically several times per day. Machine learning would have to train on the ever-changing data set, so a consumer dashboard could present the most recent and accurate information. Using crowdsourced data and the power or AI would be a game changer.
In addition, robotics process automation (RPA) could automatically aggregate and present that data in a simple, easy-to-understand dashboard that displays the individual airline performance rankings, which consumers could use to inform their air travel decisions.
There are three types of RPA technology:
- Attended, which involves continuous human oversight
- Unattended, which is autonomous, but issues alerts when errors or outliers are encountered or data values drift
- Hybrid, which has less human oversight than attended and more oversight than unattended
For airline rankings a hybrid approach is initially preferable because an RPA engine can manipulate data in an unattended fashion. As reliance on performance dashboards increase, airlines will want to use third-party performance data to ensure data accuracy. They will also use the data to optimize internal and customer-facing services.
Behind the scenes, inferential statistics could help verify the accuracy of the dataset and predict the likelihood of a good or bad travel experience. Using a critical mass of data from which averages and aggregated averages could be calculated, the AI engine could infer which populations will report accurately.
Airlines need to adopt new metrics to improve performance
One of the most popular measures of an airline’s “success” is its on-time score. Airline marketing and sales teams use the on-time score to reassure customers that they’re working hard to avoid flight delays that frustrate passengers. However, consumers don’t feel averages. Consumers feel variation, especially variation from an expectation. For example, an airline that claims 97% on time accuracy fails to reveal the degree of how long a passenger will wait when a flight is late.
In addition to providing consumers with timely and reliable information, airlines could use the data to improve their operations. For example, by adding metrics such as average lateness to average on time, both consumers and airlines could better navigate potential flight disruptions. This “voice of the customer” data could also make airlines accountable for their good or bad performance and enable them to pinpoint the issues customers immediately want fixed so airlines could prioritize service optimizations accordingly.
The time is ripe for AI to empower consumers – canceled flights cost customers hundreds of dollars, and baggage loss is skyrocketing.
The 2022 holiday travel season is upon us and, as always, there will be many complaints about flight delays, lost luggage and other bad experiences. Voice of the customer data could help improve airline performance and facilitate reliable decision-making for travelers. Crowdsourced data will the the key to change.