Your flight gets cancelled at 7pm. The airline app offers you Thursday. It is Monday. You are now expected to do the job of a travel agent yourself, on a phone, at a gate, with a dying battery and a boarding announcement going off over your head.
The scale of this is quietly enormous. In 2024, 102,908 US flights were cancelled and another 1.53 million arrived late, roughly one in five. Hopper's travel data arm counts 58 million seats departing US airports hit by significant disruption in 2025, up from 50 million in 2019. That is a stadium's worth of stranded people every single day.
Here is the kicker, and it is the whole business. When Hopper surveyed disrupted travellers, only 15% got their situation resolved in under 30 minutes. 43% were still working out their next move more than two hours later, and 7% never reached a resolution at all. Over half were not even told by the airline, they found out from a gate agent or by refreshing the app themselves.
And the official advice is worse than you think. The DOT's own guidance tells passengers to ask the airline to endorse your ticket to another carrier and find out yourself whether that carrier has space. The federal government's position is, roughly, go and negotiate it. Six of the ten biggest US carriers will rebook you on a competitor, but only if you ask and only on their terms.
So the market is not people who need a flight. It is people who need a flight in the next six hours, have a wedding, a pitch, a connection or a toddler in the mix, and would very happily hand the whole mess to somebody competent.
A flight rescue concierge. Roadside assistance, but for the moment your itinerary falls apart.
TAM. US carriers moved 84.1 million passengers in October 2025 alone, close to a billion a year. Against 58 million significantly disrupted seats at a $99 fee, the theoretical US ceiling is about $5.7bn a year. Nobody gets that, but it tells you the pool is not the constraint.
SAM. Take the 43% still stranded after two hours, roughly 25 million seats a year, and narrow to travellers with both a hard deadline and the means to pay. Call it 15% of that group: around 3.7 million rescuable cases a year, or a $370m serviceable fee pool before any ticket margin.
SOM. 0.5% of that is about 18,000 cases and $1.8m in fee revenue. That is a business of maybe 15 people, which is entirely reachable inside three years from one region.
ARR potential. Bottom up: an agent working a shift at a 45-minute average handle time clears roughly 10 cases, about $990 of gross fee revenue per agent per day. Ten agents across 250 operating days is around $2.5m. Add 20,000 memberships at $129 for another $2.6m, plus $25 of average margin on 18,000 replacement tickets, plus a compensation claim upsell priced anywhere near AirHelp's 35%. A realistic direct-to-consumer path lands at $5m to $8m ARR. One distribution deal with a card issuer or insurer takes it past $20m, which is roughly the trajectory Freebird was on when Capital One bought it.