Operational crisis and hitting rock bottom
In 1995, Continental Airlines consistently ranked at the very bottom of every performance metric in the airline industry, from chronic flight delays and cancellations to losing thousands of passenger bags each month.
The situation reached its peak when the airline became a national joke after being featured on David Letterman’s “Top 10 List” with the line: “The airline no basketball player wants to fly!” At the time, basketball was America’s most popular sport. Given the show’s massive reach, Continental was publicly humiliated before more than 5 million viewers, becoming the poster child for a brand “so bad that nobody wants to use it.”
Remarkably, just a few years later, the airline staged a complete turnaround, becoming one of America’s most beloved brands. What drove this extraordinary transformation? Let’s explore Continental’s actions below.
The big challenge: Reversing customer sentiment
With the goal of going from worst to first, CEO Gordon Bethune launched a comprehensive service overhaul focusing on cleanliness, safety, and reliability. He also introduced a new policy that rewarded each employee $100/month if planes landed safely and on time. With this new policy, within just one month Continental Airlines rose to the top of all on-time performance rankings.
But Continental didn’t stop there. They wanted to go from first to favorite. To achieve this, they needed an idea powerful enough to create a “reversal effect” in customer perception: from anger and resentment to positive emotional connection. They needed customers to genuinely feel the brand’s efforts to change and improve, thereby building real trust and positive emotions.
However, the information needed for this strategic decision was quite murky. At the time, Continental had 45 datasets, but they were scattered throughout the organization. Consolidating and connecting all this data into a single centralized source for analysis and insight discovery was a massive challenge in terms of both budget and time.
The marketing team faced two challenges:
1. Develop ideas to increase customer affinity, reversing emotions from anger and resentment to love, even turning customers into positive advocates for the airline.
2. Fragmented data (silos) and time pressure
One letter and two datasets: A dramatic shift in customer response
The marketing team proposed a remarkably simple idea: a personalized apology letter. The letter would be sent within 12 hours of incidents such as flight cancellations or lost luggage, with a sincere apology accompanied by one of two small gifts:
- A frequent flyer miles bonus, or
- A complimentary Presidents Club lounge membership at major airports.
What made this campaign different was: personalization based on customer information, transaction history, and specific incident details. The campaign didn’t cost much, but it required accurate data and fast action. So instead of waiting to consolidate all 45 data silos, they tested with just two readily available data sources: customer profitability data and service incident data.
The airline designed a test with three groups for comparison:
- A group that received no letter
- Letter recipients split into two sub-groups:
- Group receiving a letter with a frequent flyer miles bonus
- Group receiving a letter with a complimentary Presidents Club lounge pass
To evaluate the early-stage effectiveness of the test, Continental chose focus groups for their ability to deeply explore customer insights and feedback at low cost, with simple and quick implementation. During these focus groups, two completely different response patterns consistently emerged:
- No-letter group: Most customers shared their terrible experiences caused by lost luggage and flight delays. Discussions quickly turned negative, spreading feelings of hatred toward Continental Airlines.
- Letter group: Customers shared their surprise and gratitude. The apology letter was already waiting at home before they’d even finished their trip (paper mail was still common then): “I’ve never had a company apologize to me so sincerely.”
In the early stage, qualitative evaluation from focus groups provided confirming signals about positive customer responses to personalized apology letters. This evidence was enough to support the idea and allow the airline to expand the campaign more broadly.
These positive customer emotions eventually translated into clear business results as many customers converted to paid Presidents Club memberships after their complimentary trial. And when Continental completed the integration of all 45 data silos and could calculate Customer Lifetime Value (CLV), they had concrete proof: customers who received the letter showed an 8% increase in CLV compared to the non-recipient group.
5 Key Takeaways
Although this campaign took place over 30 years ago, it remains a classic case study for applying data to marketing because of its simple yet effective principles:
- Don’t wait for the perfect system: Despite having 45 fragmented data systems, Continental didn’t wait until the infrastructure was complete before starting. They began immediately with 2 valuable available datasets, combined with A/B testing and qualitative research through focus groups. This is a textbook example of “agile analytics.”
- Improving your product is NECESSARY, but making customers feel that improvement is SUFFICIENT: Capturing and responding promptly to customer pain points helps brands build lasting trust. And meaningful action isn’t just fixing errors. It’s sharing and acknowledging customer frustration, building deeper connections. In this case, Continental didn’t just improve their service; they went further with sincere apologies, emotional empathy, and a desire to make things right.
- Data isn’t just numbers: When people think of data, many marketers still think of numbers and dashboards. But in reality, data is everything that can provide information for decision-making: numbers, text, images, qualitative surveys, and more. In Continental’s case, they combined quantitative (CLV models) and qualitative (focus groups) methods for a comprehensive view.
- Test small, scale after success: Start with a small group to validate, then expand to larger systems. This is how to “fail fast and fail small,” finding a successful model before scaling up, minimizing damage if the idea doesn’t work in practice.
- Data fragmentation needs to be addressed as soon as possible: It doesn’t matter how much data you have. If you can’t use it and transform it into actionable information, that data is worthless. A/B testing combined with focus groups helped Continental find early signals to evaluate their idea’s effectiveness before scaling. But only when they consolidated all 45 data silos could they calculate critical metrics like CLV and clearly quantify their strategy’s impact.
What could we do better if this case study happened today?
Since this case study is quite old, here are some thoughts on what could be done differently with today’s technology:
- Response time: With current technology, post-incident emails could be sent within minutes, ensuring they arrive right when the customer finishes their flight.
- Enhanced personalization: Beyond just names and flight history, AI and the vast amounts of data that businesses possess today enable a new level of personalization using preferences, habits, and behavioral data to make each letter more personal and “touching.”
- Real-time analysis and engagement: Combining flight data, customer data, and real-time social media sentiment analysis to respond immediately.
- Creating advocacy: Encouraging customers who’ve made the journey from extremely dissatisfied to satisfied to share their stories on social media, amplifying the positive impact of the brand’s improvement efforts.
Beyond reversing customer emotions from resentment to affection and increasing revenue, this turnaround also laid the foundation for Continental’s marketing team to develop a long-term data-driven approach.
See you at the next “From Insights To Impacts”!
Case study source:
Jeffery, M. (2010). Where Do You Start? In Data-Driven Marketing (pp. 31-33). John Wiley & Sons, Inc.
