I keep coming back to the airlines after Friday, November 13.
For a little over four hours that afternoon and evening, departures across the United States were held on the ground. The airlines didn’t cause it, and they couldn’t fix it. By Monday, analysts were already estimating what it had cost them, and those numbers say more about business risk than any technical report will.
What happened, without the jargon
According to Transportation Secretary Sean Duffy, the trouble started during planned work at a remote FAA site, where a contractor was replacing an old copper line with fiber. The FAA said a router involved in the work began flooding the network with connection attempts, which destabilized aircraft communications and radar feeds. The nationwide ground stop was issued at 3:54 p.m. Eastern and lifted at 8:15 p.m.
The investigation is still open, and the final root cause hasn’t been published. For a business reader, the technical detail matters less than three facts: the work was planned, it happened on someone else’s network, and the companies that paid for it weren’t the ones doing the work.
Four hours down, a weekend to recover
Bernstein estimates that the four largest U.S. airlines, American, Delta, United and Southwest, lost about $345 million in pretax earnings, with an 80% range of $270 million to $430 million. Raymond James puts the hit to the same four carriers at $370 million to $964 million this quarter. These are analyst estimates, not reported results. The actual figures will show up in fourth-quarter earnings.
The flight numbers explain why the estimates are so large. More than 18,500 flights within, into or out of the U.S. had been canceled by Monday morning, based on FlightAware data. The stop lasted about four hours. The disruption ran through the weekend, because once aircraft and crews end up in the wrong cities, putting them back takes days, not hours.
Downtime is measured in hours. The cost is measured in days.
That applies well beyond aviation. When phones or internet go down at a business, the outage ends when the service comes back. The cost keeps running: the callbacks, the rescheduled appointments, the orders that went somewhere else, the overtime to catch up.
Most businesses sit in the airlines’ seat
The airlines didn’t run that network and didn’t schedule that upgrade. They carried the cost anyway. Most companies are in the same position with their own communications. Phones, internet, and the lines behind alarm panels, elevators and card readers all run on networks that someone else owns, maintains and upgrades.
Many of those networks are being rebuilt right now. Phone companies are retiring traditional copper landlines market by market. In October, the FCC invited comment on an AT&T application to discontinue traditional business landline service sold by its competitive affiliates in 34 states and the District of Columbia, set to be granted automatically on November 5, 2026, unless the FCC stepped in. California isn’t on that list. Separately, AT&T has applied to discontinue copper landline service in portions of 360 California wire centers, which it says affects about 15,000 business customers, on or after June 1, 2027.
Every one of those changes ends with a switch-over date, and that date is set on the provider’s timeline. The useful question isn’t whether your lines will change. It’s whether you’ll know the date and the cost before it arrives.
A backup on paper is an assumption
Two months before the ground stop, the FAA went through a smaller version of the same lesson. On September 21, the primary communications circuit at its Philadelphia approach control facility failed. When the system tried to switch to the backup line, the agency found that a construction crew had cut that backup fiber earlier the same morning. Flights into Philadelphia, Newark and the New York airports were halted for hours. According to the Associated Press, the circuit that failed was already due to be replaced as part of the air traffic control overhaul.
Two details in that story apply to almost any business. The risk was known: the part that failed was already on the replacement list. And the backup existed, but its condition only became clear at the moment it was needed. A backup line that isn’t monitored and tested is closer to an assumption than to protection.
Put a number on one hour
Most technology costs get discussed as a monthly bill. The cost of downtime is easy to underweight when it has no number attached. For a reference point: in Uptime Institute’s Annual Outage Analysis 2026, 57% of respondents to its 2025 survey said their most recent major outage cost more than $100,000, and one in five said more than $1 million. Uptime’s research focuses on IT and data center outages, so treat that as a benchmark from larger operations, not a forecast for yours.
Your own number is more useful, and a rough version fits into one meeting with your finance and operations leads:
- Revenue that depends on phones or internet in a typical hour: orders, bookings, payments, service calls.
- Payroll for the people who can’t work while systems are down, plus the overtime to catch up afterward.
- Missed calls and appointments, and how many of those customers come back.
- Contract exposure: service credits you owe customers, or penalties in leases and agreements.
- Recovery time: not when the line comes back, but when the business is back to normal.
Multiply the hourly figure by a realistic recovery time, not just the length of the outage. That’s the number to weigh against the cost of a second connection, a tested backup, or a better-planned switch-over.
Five questions before the next change
- What does one hour without phones or internet cost us, in dollars?
- How long would it take us to fully recover, not just to reconnect?
- Which of our lines are on a provider’s retirement schedule, and what are the dates?
- When a provider changes something on its side, who finds out first: us or our customers?
- When did we last switch to our backup on purpose, to see whether it works?
The airlines didn’t make the change. They paid for it. The difference for most businesses is that their next change hasn’t happened yet, so there’s still time to put a number on it.
How we help
We are an independent technology advisor, working on your side of the table. We help you map which phone and internet services your locations depend on, what is scheduled to change, and what an outage would cost, then compare providers for the replacement or the backup in plain business terms. If a change like this is on your calendar, let’s talk.
Sources
- Flightradar24, “US nationwide ground stop lifted,” November 14, 2026 (ground stop issued 20:54 UTC, lifted 01:15 UTC)
- Associated Press via OPB, November 16, 2026 (cause as described by the Transportation Secretary and the FAA)
- The Register, “FAA blames ground stop on faulty copper wiring,” November 2026 (cause still under investigation)
- Reuters via The Detroit News, “FAA ground stop to cost top four airlines $345 million, brokerage says,” November 16, 2026 (Bernstein estimate; 18,500+ cancellations based on FlightAware data)
- Quartz, November 16, 2026 (Bernstein 80% range, $270 million to $430 million)
- Skift, November 16, 2026 (Raymond James range, $370 million to $964 million)
- FCC Public Notice DA 26-1075, October 5, 2026 (AT&T business landline discontinuance, 34 states and D.C.)
- California Public Utilities Commission, “Information on AT&T’s Federal Applications to Discontinue Landline Voice Service Throughout California”
- Associated Press, “Flights disrupted across US Northeast due to technical problems at key air traffic center,” September 21, 2026
- Airways, September 2026 (timing of the backup fiber cut)
- Uptime Institute, Annual Outage Analysis 2026


