How Airlines Get Customer Experience So Wrong with So Much Data

Using data to segment products and customers may not breed airline satisfaction

Peter Fader flies a lot.

The Wharton School marketing professor has the highest frequent-flyer status on a certain airline, often earning him first-class seating, but he flew coach a few summers ago with his family. This discrepancy in his typical seating data must have triggered an alert to the flight crew. Once all passengers were seated, one of the pilots came back and greeted Fader, remarking on what a valuable customer he was.

“We know that you usually fly first class with us, and deservedly so,” Fader recalls the pilot saying. “But we appreciate having your whole family with you and want to make your flight as special as possible, despite being in coach.”

His children seemed impressed with their father’s status on the airline, but Fader’s mind quickly turned to the passengers around him.

“This is just weird,” he thought. “Don’t embarrass me.”

A gesture intended to preen one customer’s feathers may have instead ruffled those of others. Fader’s brief and otherwise innocent encounter with the pilot illustrates a marketing paradox: reams of customer data may not yield high customer service scores.

Airlines collect an incredible amount of passenger data, but it hasn’t helped the industry’s notoriously mediocre customer service scores. This conflicts with the marketing concept that more customer data creates more personalized experiences and greater satisfaction. Headlines decrying violently deplaned passengers suggest 2017 was a particularly chafing year for airline passengers, with customers forced to meet puritanical dress codes before boarding, removed from overbooked flights and squashed into ever-shrinking seats.

If airlines tuned out the siren song of customer data and segmentation, they would see that flying isn’t an individual experience, it’s communal. It stands to reason that members of a flight can only be as happy as the least happy customer. While it’s important to make the most valuable passengers’ experiences the best possible, airlines need to use data to lift morale for all customers, rather than a select few.

When Bad Analytics Happen to Good Passengers

Fader’s personal experience shares a thread with that of United Airlines passenger David Dao, who was physically removed from an overbooked flight last year. Both interactions were based on customer data.

Which passenger gets bumped from an overbooked flight is an automated decision, according to United’s report on Dao’s viral incident, and loyalty and money factor into the decision. Passengers who pay the least for their ticket are first in line to be bumped involuntarily, and those in United’s MileagePlus frequent-flyer program aren’t bumped unless everyone on the plane has status, in which case those with the lowest status get bumped first.

The airline industry may be the only place where segmentation determines whether customers receive the service they paid for. Imagine purchasing a meal at a restaurant but being denied that food because there isn’t enough for everyone. The more loyal customers receive the meal; you have to come back and eat at a less-convenient hour.

It’s not difficult to understand why airlines focus on their most valuable customers’ data and experience. According to Investopedia, business travelers make up 12% of passengers, but they are typically twice as profitable for airlines. Business passengers sometimes represent 75% of an airline’s profits. First-class and business tickets can cost as much as 10 times the price of coach tickets. Luring and keeping business customers makes more economic sense than focusing equally on all passengers—or so it would seem.

Dao’s forced removal, justified by the company’s algorithm that identified him as one of the least valuable people on the flight, highlighted the negative cost of such competitive customer segmentation. United Airlines’ social sentiment fell almost 160% within 48 hours of the incident, according to Brandwatch, and United Continental Holdings’ stock was off about 4% in the immediate aftermath, knocking close to $1 billion off the company’s market value.

Yet airlines continue to focus on how they can slash conveniences for those unwilling to pay a premium. Data support this strategy. A Reuters/Ipsos opinion poll released in mid-2017 found 83% of Americans put ticket prices among their top considerations when booking personal travel, outweighing travel perks or airline reputation. Sixty percent said they would not pay extra to avoid being assigned a middle seat, and 52% said they would not pay more to fly on their preferred airline. 

Armed with this knowledge, airlines have unbundled the services that were previously included when purchasing a single ticket. Booking a flight now means choosing from an a la carte group of products. Carrying on luggage, extra leg room and in-flight entertainment are now considered extras on some flights.

“Airlines have gotten pretty refined in their analysis of how much they can charge for add-ons, whether it’s checked baggage, overhead bins, advanced boarding, preferred seat selection and so on,” says John Strong, an aviation expert at the College of William and Mary’s business school.

This phenomenon has been referred to as “calculated misery,” a term coined by Tim Wu, an author and professor at Columbia Law School. In 2014, Wu wrote in The New Yorker about the decline in quality of commercial airline service, explaining: “For fees to work, there needs be something worth paying to avoid. That necessitates, at some level, a strategy that can be described as ‘calculated misery.’ Basic service, without fees, must be sufficiently degraded in order to make people want to pay to escape it. And that’s where the suffering begins.”

Strong says customer data show that airlines can take their cheapest ticket and add or subtract some nuances because people are willing to pay or save a bit, whether on leg room, reserved seating or baggage. “What data has let [airlines] do is divide economy seating into three classes now: premium economy, economy and economy-minus,” he says.

The concept of calculated misery is exemplified by leg room. The shrinking airline seat is well-documented. The New York Times reported that seat width at the time of airline deregulation in the 1970s was 18 inches, compared with about 16.5 inches today. Seat pitch, or the distance between a point on one seat and the same point on the seat in front of it, fell from 35 inches to 31 inches. Meanwhile, the average American has grown in size.

At a time when many marketers are aiming to delight their customers, airlines appear to be doing the opposite. These companies are instead marketing their products as a reduction to discomfort, instead of incremental improvements to baseline quality.

What (and from Whom) They’re Gathering

The unbundling of flight products and online booking presents airlines with an abundance of customer data. For the occasional traveler, the airline may be able to track credit cards used, flight frequency, typical destinations and spending behaviors.

More valuable customers—frequent flyers and high-status loyalty members—are tracked with a close eye. They may receive a personal apology from a gate attendee if he notices the passenger’s previous flight was delayed. A flight attendant may offer the flyer his or her favorite beverage once the flight is in progress.

These gestures, typically reserved for high-value customers, are becoming more common as airlines upgrade their technology and combine data touch points. For example, Delta Air Lines introduced software last spring called SkyPro, which may be accessed on its flight attendants’ Nokia Lumia mobile devices to track basic customer information. American Airlines is expected to launch iSolve this year, an app on its flight attendants’ Samsung Galaxy Note devices that will allow them to offer frequent-flyer miles or a travel voucher as immediate compensation for customer service issues onboard.

Rohit Deshpande, a professor of marketing at Harvard Business School, has authored two case studies on Singapore Airlines in the past 10 years. The company is considered the gold standard in customer experience, not just in its own industry, but in the greater hospitality sector, Deshpande says. Singapore Airlines worked with Accenture to develop an app that can track customer data from check-in at the airport through the flight and on to baggage retrieval.

“Let’s say you go from Sydney to Singapore to change flights from Singapore to London,” Deshpande says. “You decide when you’re going to Singapore that instead of drinking your normal gin and tonic, you’re going to ask for a vodka and tonic. They will record that, and when you reboard in Singapore to go on to London, the flight attendant will ask if you will continue with vodka and tonic or go back to gin and tonic. This might seem intrusive, but their feeling is that people want to know that they’re cared for. They want to be recognized with not just customer service but personalized customer service.”

This isn’t the only high-tech way airlines are collecting customer data. Delta has used heart rate monitors on volunteer customers to track their heartbeats at 11 stressful moments during the travel experience, such as finding a parking spot at the airport, moving through security and boarding the plane.

“We do focus groups with biometrics and biodata on where customers are looking at things on screens, where their eyes are on the airplanes to make things more intuitive,” says Andrew Wingrove, Delta’s managing director of product and customer experience. “The heart monitor was one of the tests we used for focus groups to better understand where their anxiety is.”

The monitoring extends to the web via social media, which offers a real-time opportunity for airlines to react to customers, regardless of their status.

“We’ve created a robust social listening platform to react in the moment when things are going right or going wrong,” Wingrove says. “[Customers] use it to reach out directly or to mention us. Our goal is to be where customers are, and while that may have been solely over the phone in the past, we’re thinking about it as more engagement with multiple channels.”

JetBlue Airways is one of the top brands for customer service on social media. Danny Cox, director of customer support and insights at JetBlue, acknowledges that much of the social media interaction the brand has with its audience is fun banter that doesn’t necessarily turn into data, but it’s still an important way to gather information.

“[Social media is] sometimes the first platform where people speak up,” Cox says. “I like to be at the crossroads of the very micro customer feedback, which is the single tweet, the single e-mail and then the single phone call, but also at the macro [level] where we roll out these million-plus surveys a year.”

Both Cox and Wingrove espoused the virtues of their surveys, perhaps their least tech-savvy form of data collection. Many airlines send passengers a survey post-flight to ask about their experiences at various touch points.

The JetBlue survey, which is only sent to one-third of the passengers to avoid communication fatigue, asks passengers what made them choose JetBlue, if they would recommend it to others, how they booked the flight, how the preboarding and boarding experiences ranked, if they self-tagged their luggage at one of the new kiosks and how they interacted with flight attendants among many others. More questions are asked than most survey takers respond to.

“We understand that someone may want to start a survey and not end it, but once they’ve started a survey and completed just part of the process, we’re able to collect the data up to what they had desired to complete,” Cox says.

Therein lies the first issue with surveys: If many do not complete them, perhaps only problems identified at the beginning of the questionnaire are corrected. The second issue is a matter of when the airline delivers the survey post-flight. If customers don’t receive the survey e-mail promptly, they may forget the details of their experience. The third problem is emotional.

“The research on survey response is that it tends to be answered by people who are really happy or people who are really upset,” says Deshpande. “Typically, you get fewer of the people who are really happy. Six times as many people who had an unhappy experience will provide feedback than people who have a happy experience. I don’t take great comfort from feedback surveys, especially in the hospitality industry.”

These data pose the question of whether surveys put the onus of satisfaction on the customer: Provide insights, or risk having a poor experience in the future. Customers clearly aren’t thinking this way, as The Wall Street Journal reported in 2013 that many broad e-mail questionnaires get response rates under 5%. At the time, in 2013, United said it received 10% of its blast surveys back; JetBlue said 15% to 20% of customers responded; American said its selective surveying earned an 18% response rate and Virgin America saw a 20% response.

JetBlue, for one, is acknowledging it may reach a point where passengers won’t fill out its surveys.

“I continually tell my team, ‘Imagine a day where nobody filled out a survey for us,’” Cox says. He says that when customers put trust in the company by downloading their app, for instance, the data JetBlue can collect should be insight enough. “With millennials becoming a more prominent part of our customer base, and those coming after, they’re not going to fill out surveys. They’ll think, ‘You know when my plane got there, you know when my bag got to the carousel versus when I got there. You know how that would make me feel.’”

Perhaps a more telling survey comes by way of the American Customer Service Index. In 2017, U.S. airlines as an industry drew a score of 75 on the 100-point scoring system. It’s the best score the nation’s airlines have ever received, but still keeps them in the bottom third of more than 40 sectors the ACSI measures in the U.S. economy. The authors of the report summary note that “much of the increase in passenger satisfaction appears to be driven by price,” which isn’t commonly considered a key indicator of customer satisfaction.

“Customer satisfaction has never appeared to be a goal for airlines,” ACSI Chairman Claes Fornell said in a press release. “Compared to other industries, the financial return on passenger satisfaction is not much of an incentive. The exception is in the few airports where airlines actually compete with one another—or when they treat passengers spectacularly badly in public.”

Airlines have plenty of data, but they don’t always use it proactively to enhance customer experience. Rather, it’s used to make a reactionary decision. Using United Airlines as an example, Strong explains that before the Dao incident, the discussion about overbooked flights largely centered on who would and wouldn’t get bumped, instead of reimagining the bumping process altogether.

“What [the airlines] realized upon doing customer research after those terrible experiences was that if we just make it easier to volunteer and we pay some more money, we can solve a lot of problems before they become problems,” Strong says. The experience proves Fornell’s theory that the financial return on customer satisfaction is sometimes only incentive for airlines when there’s bad PR.

Strong suggests that the culture of how data is viewed and used is the root of the problem. Airlines consider their business a logistics operation, not a customer operation.

“[Operational issues] have been the ethos of the airlines for a long time,” Strong says. The marketing and customer service teams are doing a good job of listening to customer surveys after the flight ends or using data to identify and attract customers, he says, but real-time responsiveness is lacking.

The employees should be more empowered with customer data. In particular, he says, data should help mitigate problems before they arise.

Using Data to Unite, Rather Than Divide

The divisions between ticket classes, problem solving and customer satisfaction are particularly glaring in a space where everyone is watching. Coach passengers must trudge past first-class amenities as they take their seats. The entire plane can witness a single passenger’s poor treatment by staff (along with many on the internet, if the incident is recorded). Hundreds or thousands may learn about a lost piece of luggage if the passenger takes to social media to complain.

In a communal setting, airlines and their passengers would be better off if customer analytics were used to first improve everyone’s experience. Consider the two highest-ranking U.S. airlines on the ACSI list: JetBlue and Southwest.The latter has banked its reputation on a lack of segmentation. There are limited options or products when purchasing a Southwest ticket: Business Select, Anytime or Wanna Get Away. The only perks to the most-expensive tickets are refunds, a free “premium drink,” more Rapid Rewards points and priority boarding. Southwest planes don’t have visible passenger seating distinctions. The only notable add-on is EarlyBird Check-In.

“Southwest has decided that even though they know there are lots of people who will pick and choose whether they want this feature or that feature, there’s also a customer segment that just wants a consistent product,” Strong says.

The low-cost airline’s research dictated the decision not to break out separate products or fees. Paul Sacco, Southwest’s director of customer intelligence and strategy, says the company chose not to follow the industry trend of charging for bags because its customer data showed both infrequent leisure travelers and its most-frequent business travelers used the checked bag service and appreciated the free benefit.

Sacco acknowledges that personalization has become a customer expectation, but maintains that the core service needs to back that up. The airline’s research on its customers speaks to the importance of both well-functioning logistics and positive customer service.

“We have found that on-time performance and positive interactions with our customer-facing employees are the strongest drivers and predictors of customer satisfaction and customer likelihood to recommend Southwest,” Sacco says.

Southwest received an ACSI score of 80, ranking second on the list among U.S. airlines. The company’s other metrics, such as canceled flights and on-time arrivals tracked by the Department of Transportation, are less impressive. It suggests the work Southwest has done in transparent pricing, egalitarian seating and pleasant customer service simply matters more to its customers.

The customer service aspect is particularly compelling. Southwest gates in some airports in December were impressively decorated with handmade snowflakes and old-fashioned sleds, a stark contrast to the sparsity of other gates. And in the era of viral videos, it’s also worth noting that the top four clips that appear on YouTube when searching “best flight attendant” are of Southwest employees.

JetBlue topped the ACSI airline rankings—besting Southwest by two points—by striking a balance between a positive baseline experience and added perks. The building blocks have been customer data and analytics.

For example, JetBlue noticed that it had consistently negative trends in Philadelphia. A review of both logistical data and customer reviews showed there was a timing issue: Flights weren’t late, but passengers were arriving at an hour that coffee shops in the airport weren’t yet open.

“All of these negative [reviews] were because people weren’t getting their cup of joe before jumping on this flight from Philadelphia to their meetings and feeling prepped and primed for that,” Cox says. “We could have just stopped at, well, people in Philadelphia, they’re just angry. But, instead, we dug deeper to see what was really needed. We could easily partner with the coffee shops there and ask them to adjust their hours. Immediately the scores popped back up to a more reasonable and expected area.”

Cox says as customer expectations have risen, JetBlue has worked to keep up while maintaining the base. The airline may be known by budget-conscious consumers who follow JetBlue’s flash sales, but it also offers products like Mint, its premium service, as an upgrade.

Like Southwest, the airline doesn’t rank particularly well on traditional metrics (2017 was JetBlue’s worst year for flight delays since 2007), so price and customer service once again outrank other factors. Being on-time certainly matters, but Strong’s observation that airlines focus too much on logistics appears to be true. United ranked highest for fewest canceled flights and second for most on-time arrivals, but ranked toward the bottom on customer satisfaction scores. It’s a prime example of why airlines need to shift their data focus to customer service, but they also need to understand their actions are in full view by an entire airplane of customers.

There’s an obsession with customer data and personalization right now. The theory is that individualized experiences breed loyalty; but here’s a caveat: Most of these experiences don’t happen with everyone else watching, or at others’ expense. When Target’s “pregnancy prediction” score determines a customer is expecting, it doesn’t offer deals on baby-related items in full view of childless couples. When Starbucks customers join the loyalty program, they don’t get to brandish their cards to cut in front of non-members in line.

Airline customers may each be unique, but they have a communal experience when flying. The surveys, formal and informal, suggest everyone benefits when the general population is content and the loyalists are (discreetly) acknowledged. No matter where a person sits or how much they pay, every passenger feels the same turbulence.