Few industries lose as many ready-to-buy customers as travel. Around 81 percent of all started bookings are abandoned — for airlines, the figure is nearly 88 percent. The paradox: purchase intent is there, and it's higher than in retail. Yet the booking funnel loses those customers anyway. Why booking processes fail, why abandonment doesn't mean disinterest, and how personalization turns browsers into bookers.
Customers want to buy. The funnel won't let them.
The travel industry has a problem that looks like a luxury problem at first glance. Interest is enormous; completed purchases are rare. The vendor SaleCycle has published what is likely the largest analysis of its kind in the industry. The dataset covers more than 280 million online bookings. The finding: around 81 percent of all started bookings are abandoned. For airlines, the rate is as high as 87.8 percent.
To put that number in perspective, a comparison with retail helps. In fashion, about 11.5 percent of visitors add something to their cart. In travel, nearly 18 percent of visitors start a booking. Purchase intent is significantly higher than in retail — it just never reaches the finish line. Put differently: marketing has done its job. Customers are there, searching, browsing, beginning to book. They're lost in the final stretch.
That's exactly what makes this number so valuable. It shows that the biggest revenue opportunity for most travel providers isn't more traffic. It's the visitors who were already there.
Why booking abandonment is especially high in travel
There are understandable reasons why abandonment rates in travel are higher than almost anywhere else. Travel is expensive. Someone spending €2,000 on a family holiday makes decisions differently from someone ordering a T-shirt. They compare, sleep on it, consult their partner. Some abandonment is a natural part of the decision process and will never go away entirely.
But there are also self-inflicted causes. Prices that change mid-booking and erode trust. Forms asking for passport numbers and frequent flyer accounts before the customer has even made up their mind. Fees that appear only in the final step. And booking flows that show every visitor exactly the same thing — a first-time visitor the same experience as a customer who has searched the same route five times.
The first category of causes you cannot eliminate. The second you absolutely can. And the second is larger than most providers believe.
The misconception: abandonment doesn't mean disinterest
In many booking systems, an abandoned session simply disappears. No signal, no response, no follow-up. That reflects a fundamental misconception. Someone who abandons a booking has, in the vast majority of cases, not lost interest. They have a question the funnel didn't answer. Or they simply need more time.
To make this concrete: a customer searches the same flight to Lisbon for the third time on a Tuesday evening. That's not wasted traffic or a lapsed lead. That's someone on the verge of a decision. Most systems still show her the standard homepage, as if she'd never been there. She has to type her search again, filter again, compare again. Every one of those steps is a chance for her to give up.
The right response would be straightforward. When she returns, she sees her last search and the connection she looked at — right at the top. Maybe with a note on whether the price still applies. That's not rocket science. It's common courtesy, expressed in software.
What personalization along the booking funnel looks like with Dynamic Yield
Personalization sounds like a major undertaking. In a booking funnel, it comes down to many small, very specific decisions. A few examples illustrate the pattern.
First-time visitors need orientation. They get the most popular destinations, clear entry points, and trust signals. Return visitors need continuity. They get their last search and saved items — not the standard experience. A family searching during school holidays sees different content than a solo traveler looking for a two-day trip to Berlin on a weekday. Someone hesitating at the final step gets answers to the questions that are typically open at that point: what does cancellation cost, what's the baggage policy, when will the charge be taken. Clearly visible — not buried in a help section. What they don't need is another banner with an upsell offer.
None of this requires magical data. The signals are already there: search history, travel dates, device, time of day, traffic source, behavior in the current session. The only question is whether the booking funnel uses those signals or ignores them.
A/B testing with Dynamic Yield: measure, don't assume
A fair objection arises here. How do you know which of these measures actually works for your customers? The honest answer: you don't — not in advance. That's why every personalization effort needs a testing program. Every measure runs as an experiment first. A portion of visitors sees the new variant; the rest see the original. Success is measured against real bookings, not clicks.
The model for this comes from within the travel industry itself. Booking.com runs more than 25,000 experiments per year and has built one of the world's strongest conversion engines through this approach. Getting there wasn't the result of one big initiative — it was thousands of small improvements, each proven to work. We explored why this culture matters more than any single feature in the article Schluss mit Bauchgefühl.
Technically, both disciplines now run on a single platform. With Dynamic Yield, you manage personalization based on behavior and context while simultaneously testing every variant across your website, app, and email channel. For a travel provider, that means: form a hypothesis, measure it against real bookings, roll out what demonstrably reduces abandonment. Then start again.
Three starting points for travel providers over the next 90 days
Where do you begin when your own abandonment rate is somewhere around 80 percent? In our view, with three steps that build on each other.
The first step is an honest measurement of your own funnel. Where exactly are you losing customers? During search, during selection, in the data-entry screen, at payment? Many providers know their overall rate but not where the drop-off actually happens. Without that data, every measure is a shot in the dark. With it, a vague problem becomes a concrete area to fix.
The second step is the return experience. Build the experience for returning visitors with a known search. This is your warmest segment — and almost always the most poorly served. Show the last search again, surface the connection they viewed, communicate the current price. This single use case typically offers the best ratio of effort to impact.
The third step is a testing program with clear rules. One use case, one testable hypothesis, one result that counts. You'll learn more about your customers from the first test than from any workshop. Everything else follows from what you discover.
The 81 percent is not a law of nature
Some abandonment is part of the travel business and always will be. People compare, hesitate, and decide together. But the gap between an average booking funnel and a good one is large — and measurable. Every percentage point reduction in abandonment is revenue from visitors you've already paid to acquire.
The 81 percent is therefore not a threat, but an invitation. It describes the distance between what your funnel shows today and what your customers actually need. That distance can be closed — step by step, test by test. Find out more in our Dynamic Yield Insights.