Exactly How the Algorithm Was Working Against Him
Large hotel booking companies need to make money. That part is completely reasonable.
They built the infrastructure. Negotiated with thousands of hotels across hundreds of countries. They run the servers, process the payments, handle the disputes. A platform that does all of that is entitled to charge for it. Nobody serious argues otherwise.
What they do with the search results, though – that’s a different conversation.
What the Algorithm Actually Does
When a traveler searches for a hotel, they assume the results are ordered by relevance. Best match first. Most value at the top. The platform working on their behalf.
That assumption is mostly wrong.
Results are ordered by what’s most profitable for the platform. Hotels that pay higher commission surface higher. Sponsored listings appear at the top dressed as organic results. A perfectly good hotel two blocks from where the traveler needs to be sits on page three because it didn’t bid high enough. The traveler scrolls past it without ever knowing it existed.
But here’s where it gets interesting.
Search for a hotel. Spend a few minutes on one listing – clicking through the photos, reading the reviews, checking the room types. Close the tab. Come back two hours later.
That hotel you spent the most time on? The price just went up.
Not every hotel. Not the ones you glanced at. The one you lingered on. The one the platform now knows you want.
Accident? The engineers who build these systems would find that question very amusing.
The platform isn’t lying exactly. It’s just quietly answering a different question than the one the traveler asked. It knows what the traveler wants. It’s deciding how much that want is worth – and pricing accordingly.
He knew this because building recommendation systems was his job.
City Twelve
He works at a sizeable AI company – not saying which one, though there’s a reasonable chance the name would be familiar. Travels constantly. Conferences, client dinners, last-minute meetings in cities that weren’t on the schedule at the start of the week.
He started counting at some point. Nineteen cities in a single quarter. Somewhere around city twelve – Seoul, Tuesday night, dinner that ran three hours longer than planned – he booked a hotel two streets away without thinking. No research. No comparison. Just opened an app, picked the first result, paid, slept, left.
His colleague watched from across the lobby and said: you’re not traveling anymore. You’re just crashing.
The word meant more than intended.
Because crashing – finding somewhere to land without the ceremony, paying only what the room is worth, spending the rest on the reason for the trip – was the most honest description of what frequent unplanned travel actually looks like. And nobody had built a booking tool that served that traveler honestly.
The big platforms were built for the vacation planner. The itinerary maker. The person with three weeks of lead time and a Pinterest board.
Not for the person landing at 10pm needing somewhere decent before 7am.
What He Built
He understood how recommendation algorithms decide what to show and what to bury. He understood commission structures and how they shape search results. He understood, specifically, that the first page of any major platform is less a reflection of what’s available and more a reflection of what’s been paid for.
Most importantly – he understood the pricing game. The way a platform watches how long a traveler stares at a listing and uses that attention to raise the price. The way hesitation gets monetized. The way the algorithm meant to help find the best room is actually measuring how badly the traveler wants it.
So he built something that worked the other way around.
Type “near the venue, quiet, available tonight” or “beachfront, dog friendly, nothing over $90” and the AI reads the intent, cuts through the sponsored layer, and finds what the market genuinely offers – before the interest detection kicks in and the price moves.
The commission: one percent. Not fifteen. Not thirty. One.
He called it Crashers – because that’s the traveler it was built for. The one who isn’t on vacation so much as crashing somewhere. Finding a decent place to land, paying only what it’s worth, spending the difference on the reason they’re there.
The .org was deliberate. Every major platform is a .com – honest shorthand for a commercial operation built around the transaction. The .org was a quiet but intentional signal. A community rather than a company. Something built around the traveler rather than the booking fee.
“Small things tell you what someone actually believes,” he said.
How It Spread
No press release. No marketing budget. No launch campaign.
The engineer shared it with colleagues, then friends, then friends of friends. It spread through group chats – Travel Hacks communities, solo traveler networks, digital nomad forums. The quiet forwarded link that arrives with one line: use this instead.
“At some point they will notice,” he said. “That’s fine. Until then it works.”
The major platforms built something genuinely impressive. The infrastructure, the inventory, the customer service, the global reach – all of them are valuable. They deserve to make money from it.
They just don’t need to make it by watching how long a traveler stares at a room and raising the price the moment they look away.
Crashers just need a comfy place to crash.
crashers.org
