The 23h problem: who answers when the trip breaks?
A business trip can be flawless for fourteen hours and ruined in the fifteenth. What happens in that fifteenth hour is the whole measure of a travel program.
It is eleven at night in a city that is not home. The flight that was supposed to leave at nine has been cancelled, not delayed, cancelled, and the board now shows the next option leaving at six in the morning. The gate agent has a line forty people deep and a tone of voice that has given up. Somewhere in that line is a director who has a board meeting at eight, in a different city, that cannot move. She has done everything right. She booked in policy, she arrived early, she is exactly where the system told her to be. And none of it matters now, because the trip has broken, and the question that decides how the rest of her night goes is brutally simple.
Who answers?
I have come to think of this as the 23h problem, the eleven-o’clock problem, and I think it is the truest test of a travel program that exists. Not the booking. Not the savings report. Not the slick app. Those are the easy hours, the fourteen hours when everything works and any system looks competent. The measure of a program is the fifteenth hour, the one where it all falls apart, far from home, late at night, with a real consequence waiting on the other side of the disruption.
Most of the conversation in corporate travel technology is about the easy hours. It is about the search, the comparison, the policy compliance, the dashboard. All of that matters, and I have spent years building it. But I want to argue that the easy hours are table stakes, and that the entire differentiation, the thing that actually earns a traveler’s trust and a company’s loyalty, lives in the fifteenth hour. In what happens at 23h, when the trip breaks.
Let me describe what usually happens, because most people reading this have lived some version of it.
The traveler, standing in that line, pulls out their phone. They open the app they booked through. And here is the first fork in the road, the one that tells you everything about how the program was designed. Either the app is a tool that hands the problem back to the traveler, a search interface that says, in effect, “here are your options, good luck”, or there is a human, or something that behaves with the judgment of a human, already working the problem on the traveler’s behalf.
The first kind of app is the cockpit I have written about before. It is very good during the easy hours and nearly useless in the fifteenth, because the last thing a tired, stranded person at 23h needs is to become their own travel agent. They do not want options. They want the problem solved. The cognitive load of researching alternative flights, checking what is in policy, rebooking, and finding a hotel for the night is exactly the load they have the least capacity to carry at that moment. Handing it back to them, however elegantly, is a kind of abandonment dressed up as self-service.
The second kind is rarer and harder to build, and it is the only kind that actually works when it matters. It is a program where, the moment the flight cancels, someone or something is already moving. The rebooking is being worked before the traveler reaches the front of the line. The hotel for the night is being arranged. The traveler’s phone does not present a puzzle; it presents a plan. “Your nine o’clock is cancelled. We have rebooked you on the six a.m., here is your seat, we have held a room at the airport hotel tonight, and your car will be there in twenty minutes.” The difference between those two experiences is the difference between a traveler who trusts the program with their life and a traveler who quietly starts looking for a different one.
Here is the part that I think the industry has gotten backwards for years. We have treated support as a cost center, a thing to be minimized, automated away, deflected into a chatbot or an FAQ. And for the easy hours, that instinct is fine, even correct, because the easy hours genuinely do not need much human attention. But the fifteenth hour is not a cost to be minimized. It is the moment the entire relationship is decided. A company can deliver fourteen perfect hours a thousand times, and one botched fifteenth hour, one night a director spent stranded and alone with a dead app, will undo all of it. Trust in a travel program is not built in the easy hours. It is built, or destroyed, in the fifteenth.
This is where I want to be careful, because it would be easy to turn this into an argument that technology is the problem and humans are the answer. That is too simple, and it is not what I believe.
The truth is more interesting. The fifteenth hour is precisely where technology and human service have to work together, each doing what it is best at, or the whole thing fails. Technology’s job is to see the disruption coming and to start solving it before the human even knows there is a problem. A good system knows the flight is going to cancel before the gate agent announces it, because the data is already moving. It can begin the rebooking automatically, surface the options, hold the inventory. That is work no human could do at the speed required, across thousands of simultaneous travelers. The machine is extraordinary at the fifteenth hour, right up to the edge of where it stops being extraordinary.
And that edge is real. Because sometimes the fifteenth hour is not a logistics problem. Sometimes it is a human one. The traveler is not just stranded; they are frightened, or grieving, or at the end of their rope after the third thing to go wrong that day. The board meeting is not just a calendar entry; it is the most important presentation of their career. In that moment, what the traveler needs is not only an optimized rebooking. It is a human voice that understands the weight of the situation and responds to the person, not just the itinerary. The machine can hand the traveler a perfect plan and still miss that they needed, for thirty seconds, to be treated as a person who is having a hard night.
So the architecture of a good answer to the 23h problem is layered. The technology handles the speed and the scale, the seeing-it-coming and the instant rebooking, so that the vast majority of fifteenth hours resolve themselves before they become a crisis. And the human is held in reserve for the moments that are genuinely human, freed up to be present for them precisely because the machine has handled everything that did not require a person. The technology does not replace the human in the fifteenth hour. It clears the noise so the human can be there for the part that actually needs them.
This is the conviction that sits underneath how we built VOLL, and the 23h problem is, in a sense, the problem we are most obsessed with. Not because the easy hours do not matter, but because we decided early that a travel program earns its keep in the fifteenth hour, and that a program which abandons the traveler at 23h has failed no matter how good its search interface is. It is the reason the platform was built with service in its DNA rather than bolted on as an afterthought, and the reason the technology and the human support are designed to work as one thing rather than two.
But I am not writing this to talk about us. I am writing it because I think the 23h problem is the right lens for anyone evaluating how they handle travel, regardless of what tools they use. The questions it raises are simple and they cut through all the noise of features and dashboards.
When your traveler’s flight cancels at eleven at night, what actually happens? Does your program hand them a search bar, or does it hand them a plan? Is someone, or something, already working the problem before they reach the front of the line? And when the situation stops being about logistics and becomes about a person having a genuinely bad night, is there a human reachable, or does the trail end at a chatbot that loops them back to the FAQ?
Most companies have never asked these questions, because they evaluated their travel program during the easy hours, when everything works and every option looks fine. The easy hours are a bad test. They tell you nothing, because almost anything passes them. The fifteenth hour is the only test that matters, and it is the one almost no one runs before they need the answer.
The director in that line will remember this night. She will remember it far more vividly than the fourteen perfect hours that preceded it, because that is how human memory works; we remember the moment things broke and who showed up. And whatever shows up for her at 23h, the plan or the puzzle, the human voice or the dead app, will become, for her, what the travel program actually is. Everything else was just the easy hours.
The trip will break. It always, eventually, breaks. The only question that was ever worth asking is who answers when it does.
About me
I am an entrepreneur with over 20 years of experience at the intersection of tourism and technology. I am co-founder and Chief Business Officer of VOLL, the largest mobile-first corporate travel and expense management platform in Latin America, and a recognized reference in the development of the corporate travel industry.
A Marketing specialist from Fundação Dom Cabral, I serve on the Tourism Council of FecomércioSP and on the Executive Council of the Latin American Association of Corporate Events and Travel Management (Alagev). A frequent traveler and close observer of human behavior in motion, I write and speak about innovation, digital transformation, entrepreneurial leadership, and the future of corporate travel.