Henrik Zillmer and the price of giving up

Henrik Zillmer bought plane tickets for a reason the airlines probably had not put in their revenue forecasts. He wanted to get through security. On the other side were delayed passengers, and delayed passengers were potential customers. AirHelp could help them pursue compensation. First, somebody had to tell them it existed.

The tickets were cheap. The conversations were direct. Eventually, by Zillmer’s account, the airports threw them out. As an early sales strategy, it had certain limitations. As an explanation of the founder, it is rather good: find the people with the problem, get close enough to speak to them, worry about the elegant version later.

For Zillmer, the original problem had been personal. A flight from Asia arrived six hours late. Getting the money he believed he was owed took six months. The flight delay had ended; the administrative delay was just warming up. Somewhere between the two, he found the beginnings of a company.

The long way to starting

Zillmer studied International Business at Copenhagen Business School and completed an MSc in E-Business in 2009. Exchange semesters took him to Hong Kong and New York. He wanted international experience and the tools to build businesses, but collecting the tools also offered a respectable way to postpone using them.

He later admitted that he had wanted to become an entrepreneur long before he felt ready. Education and corporate work became preparation for a departure that kept moving further away. When he finally left employment, the dominant feeling was relief. “It was a total relief to leave the job,” he recalled.

His route also included comedy and service in the Danish military. Those are unusually different rehearsals for a business career. One asks you to notice when the room has stopped listening; the other gives a young person responsibility inside a hierarchy. Zillmer would eventually work in an industry where getting a response from a large organization was the central difficulty.

At Rocket Internet, he encountered a particular appetite for execution. His later account describes an environment built around speed and taking established business models into new markets. It gave him a practical education in getting things done. AirHelp would require him to apply that urgency to a problem he had experienced himself.

Three thousand messages before the machine

AirHelp’s early proposition could fit on a landing page. Enter your flight information. The company would pursue compensation and keep a share if it succeeded. In its early model, that share was 25 percent. The offer made a complicated process sound manageable, although much of the machinery behind it still consisted of its founders.

Zillmer’s co-founders were Nicolas Michaelsen and Greg Roodt. After media attention, he recalled roughly 3,000 signups in a week. For two months, the three were answering customers and working out how the service should function. The inbox became an inconveniently thorough product briefing.

There is a useful order of operations here. The founders offered an outcome before they had automated its delivery. The resulting work showed them what needed to be automated. A flight claim could be small enough to discourage conventional legal attention while still mattering considerably to the person owed the money.

AirHelp was founded in 2013 and joined Y Combinator’s Winter 2014 batch. Its early application imagined combining flight information with passenger details and notifications. The ambition was already international. The practical challenge was to make the work repeatable without treating each passenger as an entirely new administrative adventure.

An email with an unexpected ending

The Y Combinator interview supplied a neat little lesson in reading signals. Zillmer understood that a phone call would mean acceptance and an email would mean rejection. After the interview, an email arrived. It asked for his phone number. Even a startup accelerator occasionally needs somebody to complete a form.

He has written candidly about how badly he thought the interview had gone. The questions exposed weaknesses; the trip felt like an expensive mistake. Later came the call confirming that AirHelp was in. His account contains considerably more celebration than composure.

For other applicants, his advice became brisk: answer directly, keep explanations short and make sure the founders agree on what they are building. The interview is a poor place to discover that your colleague has been developing an entirely different company.

“Skip the fancy marketing talk.”

Henrik Zillmer · What Makes a Good Pitch?

He applies the same standard to a pitch. Explain the business early. Try it on your grandmother. Remove impressive adjectives that delay understanding. His impatience with ornamental language is especially appropriate for AirHelp: passengers already dealing with one incomprehensible process have little reason to welcome another.

His comparison of Rocket Internet and Y Combinator puts different emphases beside each other. Rocket sharpened execution; YC concentrated his attention on product, growth and explaining value. Speed helps you run the experiment. Clarity helps you understand what, exactly, the experiment is supposed to prove.

Small claims, many countries

Zillmer gave the broader idea a name: Justice-as-a-Service. The difficulty he described was familiar well beyond aviation. People may have consumer rights without knowing how to use them. They may understand the process and still decide that the time, correspondence and uncertainty cost more than the possible result.

His argument was that data and software could reduce the work required to assess and pursue smaller claims. AirHelp brought together flight, weather and airport information with legal processes. The economic question was whether handling each case could become inexpensive enough to make the service useful to individual passengers.

Building internationally also shaped the organization. In an early account, Zillmer listed bookkeeping in Pakistan, design in Romania, development in Poland and sales work in the Philippines. He described a company with 38 people, five freelancers and 15 nationalities. International work was already part of the daily routine.

His enthusiasm for delegation came with a qualification: founders should do the work themselves first. A brief becomes much clearer when the person writing it understands the task. He also insisted that talking to customers belonged with the founders. Some activities could cross continents; understanding why anyone bought the service required personal attention.

Payments provided another unglamorous obstacle. Helping people in different countries meant handling different currencies and payment methods. Zillmer described compensation transfers running into anti-money-laundering filters, adding cost and delay. A successful claim still needed to become money in a passenger’s account.

“Getting money from someone who doesn’t want to give it to you.”

Henrik Zillmer · describing the business in 2022

A different seat at the table

By October 2019, AirHelp had 750 employees. Zillmer announced that Jimmy Maymann would take over as CEO and that he would become chairman. Maymann had already supported the company as an investor and board chairman. The replacement was familiar with the business long before the title changed.

Zillmer’s announcement was affectionate about the company and deliberate about the handover. He wanted someone with international leadership experience beyond his own. He also planned a six-month break, travel with his fiancée and a wedding in the spring. After years of building a service for travelers, he was making time to travel himself.

Travel appears in his life as more than the origin of customer complaints. Asked in 2018 to recommend a destination, he chose Fernando de Noronha in Brazil. He named Hong Kong, New York and Thailand as places that had influenced him. The accessory that had improved his travel experience most was a pair of noise-reducing headphones. Even a passenger-rights founder appreciates a problem that can be solved by putting something over his ears.

He also recalled an emergency landing between Gdansk and Warsaw because of a flat tire. Everyone landed safely, but the experience stayed with him. It is a small reminder that a frequent traveler can know a great deal about flying and still feel vulnerable inside the aircraft. Running a company around disrupted journeys does not remove the passenger’s perspective. He had plenty of opportunities to keep it.

The decision adds a quieter complication to the airport-gate story. Early on, persistence meant doing more himself. Later, it meant trusting someone else with daily leadership. There is no automatic timetable for that change. A founder has to decide whether the habits that helped create a company are still the habits its next stage needs.

The claim that remains

In March 2025, AirHelp announced a significant minority investment from Abry Partners. It was structured as a secondary investment, allowing capital to return to existing shareholders. The announcement identified Zillmer as founder and chairman, with Tomasz Pawliszyn as CEO. His relationship with the company had continued beyond the handover.

By then, Lisbon was his base. A conversation recorded in Portugal in October 2025 and published in February 2026 found him reflecting on the many detours of entrepreneurship. He also said he wanted to write a book about the journey. There is enough administrative suspense in his acceptance email alone to justify keeping notes.

In July 2026, he revisited AirHelp’s beginnings on FlashpointSpeaks, including its shift from information toward a claims business and the role of automation. The central problem remained easy to recognize. A passenger has a disrupted journey. Pursuing a remedy takes effort. Someone has to make that effort worthwhile.

It is an appropriately practical subject for a founder who once bought a ticket simply to reach the conversation. At the departure gate, he could see the customer and explain the offer. Years later, the business still depends on finishing what that conversation started.