The Hidden Impact of Internet Failing in Disasters

Show notes

Running an ISP is never boring, and nobody knows what is around the next corner. A wildfire can take out every tower an operator owns in less than ten hours. An ice storm can coat every antenna in a service area. This episode is about what happens to an ISP, its team, and its customers when that day comes.

On this episode of Bandwidth, Georgette Lopez-Aguado, Rick Seemann, and Larry Weidig talk through planning for the unplannable: what to say and when, which calls are not worth making, how to pay for recovery, and what the work does to the people doing it.

Communication Starts Before the Storm

Georgette's first point was that post-event communication starts before the event, with a plan. An ice storm or a hurricane gives some warning. A wildfire or a tornado does not. If customers know where to look for updates and staff know where to post them, the first hours go very differently.

Once it hits, Larry's advice was to stay calm and be precise. Use your BSS/OSS to identify the customers who were actually affected and communicate with that group. Messaging everyone adds angst and confusion for people who were never impacted.

Rick added that if the outage will not be resolved by the end of the day, say so, and share the order of operations if you have one. Use every channel available: email, SMS, the website, the phone system, and social media. Running an ISP in rural Ontario, he saw how many customers did not connect the ice outside with their internet being down. A photo of a tower on the ground explains it instantly, and a recorded message on the phone line keeps support channels open for real diagnostics. As recovery moves forward, show the progress: the truck at the site, the tower being rebuilt.

The Scale of the Disaster Changes the Job

Georgette noted that a smaller tornado might be cleaned up in five to seven days, while the most severe ones mean rebuilding infrastructure for the long haul, sometimes for communities that are no longer there. In those cases the priority can shift away from restoring the traditional network and toward getting connectivity to disaster recovery sites, because that is where the community is.

She also described what she has watched operators do for each other: reaching across state lines and across countries with equipment and crews to help an impacted provider get back up. Larger incumbents show up as well.

Tone Matters as Much as Speed

Loyalty during a disaster is built before it, and it can be lost by fumbling the response. Customers want to hear from you, and they want to hear from you honestly.

Georgette shared an example of an operator that sent mass messages asking customers for details on damaged routers very early in the recovery. Larry's reaction was blunt: these people have just lost their house, their photos, and their belongings, and you are asking for a serial number on a router. That information should already be in your system. The panel read it as a symptom of having no disaster plan, which leaves a team in pure reaction mode.

For operators without a plan, Georgette suggested starting an outline now, even with an AI assistant, and filling in the blanks. Larry added that the prompts need local context, because a plan for Wisconsin looks nothing like a plan for Puerto Rico or Texas.

Insurance and Credits

Larry called insurance a math equation. His team set a dollar threshold for when a claim was worth filing, and counted the full cost of filing: pulling invoices, matching serial numbers, gathering photos. For larger outages it was worthwhile, and their policy also covered the credits issued to customers, which add up fast across thousands of accounts.

On credits, the advice was to give them proactively and generously. The outage was out of the customer's control, and Rick pointed out that making each customer ask costs money too, so one blanket adjustment in billing is usually cheaper. He also admitted his own team came to insurance late and left money unrecovered. For outages that run weeks, keep crediting, because the other risk is a competitor getting back online in that area before you do.

Safety Is Not a Tough Call

Asked about the hardest calls he made as an operator, Rick said most came down to safety. His rule: do not do anything you would not ask an employee to do. Larry was direct that people matter more than any access point or backhaul link. You cannot climb a tower carrying three inches of ice, no matter how you tie off. You wait for the sun. In his words, if it feels like a tough call, you are probably not taking the people into account.

Georgette described crews checking aerial fiber after straight-line winds, driving past downed poles that also carried live electrical lines, working alongside first responders. Deciding what counts as assessing damage and what counts as putting staff in danger is part of the job.

The non-negotiables the panel landed on were following the plan and communicating. And the plan has to be tested, not left in a binder on a shelf. Georgette, who volunteered for years with the American Red Cross in disaster recovery, recommended a post-mortem after every event: what would you do again, what would you do differently, and what would you stop doing.

Getting Creative to Get Back Online

When lightning took out more licensed backhauls than Larry's team had spares for, they turned access points into clients and fed the tower from another site. It did not run at full capacity, but customers were back online. Rick remembered improvised mounts built on the spot. Georgette has seen operators' spouses rotate through gas station lines after a hurricane to keep generators fueled, and teams set up temporary command centers when offices or inventory locations flooded.

Rick also made the case for traffic shaping during a disaster. With reduced capacity, prioritizing essential traffic makes the network far more usable, and those profiles can be prepared ahead of time as part of the plan.

The Human Side of Recovery

Georgette reminded operators that their staff are often living through the same disaster they are responding to, and that events involving loss of life mean a long recovery. She argued that a disaster plan should include mental health: know the resources in your community and make room for people to debrief.

Larry told the story of a tornado that hit a community his ISP served. He went out to check a tower at a customer's property and found a pine tree through the owner's house. At that point you stop caring about the equipment. Later, his installers put antennas on the RVs of families who had lost their homes, and the gratitude stayed with the team. Rick agreed that field technicians get to see those moments up close, and that they are a reminder of why the work matters.

What If the Power Fails?

Power failure happens in tiers, so the answer has to be layered. Larry's team moved from AC to DC-based UPS systems after the AC units caused more tower outages than they prevented. Larger sites had generators with automatic transfer switches, which only help if they are maintained and tested regularly. On shared towers, tenants were willing to split the cost of a generator. Remote power control lets you shed load as batteries drain, and portable generators are the last layer. Georgette is seeing operators deploy generators with solar panels to cut down on refueling trips.

Listen to the Full Episode

Catch the full conversation on the Bandwidth podcast.

Available now on Spotify, Apple Podcasts, and the Bandwidth YouTube Channel.

If you enjoyed this episode, be sure to subscribe wherever you get your podcasts so you never miss a conversation.