Airheart

I co-founded Airheart with Lindsey Renken. It started in the pandemic as Airheart Explorer, a plain guide to the travel rules for every country. It grew into the largest travel restrictions site on the web. At its peak, about 150,000 people a month used it. Nearly all of them found us through search. We never paid for that traffic.

When borders reopened, we turned Airheart into a flight and hotel search engine. Flight data came from Duffel. Hotel data came from Expedia, through one of the first partnerships they signed with a start-up. We then became a data provider to Google, so our fares and rates showed up in Google Flights and Google Hotels.

Google Hotels pulled live prices straight from our servers. That meant 20,000 to 25,000 requests per second, around the clock. It was the hardest scaling problem I have worked on. We spent about US$250,000 in cloud credits to keep it running for a year.

In the end, we shut it down. Three things made it unviable:

• Fraud. Card fraud in online travel is relentless, and a small team pays for every stolen card.
• Thin margins. Rate parity rules mean every site must show the same price. You cannot compete on price.
• The supplier is the competitor. Expedia owned the inventory, set the terms, and had the brand. Whatever margin was left, they took.

My takeaway: a start-up cannot win in travel booking by reselling someone else’s inventory. The engineering was the fun part. The business was never going to work.

My friend Fenn Bailey learned this long before I did. He co-founded Adioso, a flight search start-up, in the same Y Combinator batch as Airbnb in 2009. His advice to anyone who asks: never start a travel business.

One of our investors went further. In their view, every consumer product ends up as an arbitrage play. Consumers will not pay enough to fund the product, so you end up selling their data or selling ads.