Everything works fine until one person takes a day off.
That line hit harder than it should, because if you’ve been anywhere near a growing ISP, you’ve lived it. Things hum along, installs are getting done, customers are getting connected, and then suddenly something small breaks. Not because the network failed, but because a process did.
That’s what this episode really gets into. Growth doesn’t just bring more customers. It exposes every gap you didn’t know you had.
When you’re starting out, you’re doing whatever it takes to get customers online. Manual processes are fine. Tribal knowledge is fine. Someone knows the answer, and if they don’t, they figure it out.
But those same habits don’t scale.
You can get away with spreadsheets for inventory or IP assignments for a while. You can route installs based on who’s available instead of what’s efficient. You can even run support off someone’s cell phone in the early days.
Then one day, you can’t.
You don’t gradually notice the shift either. It shows up all at once. You wake up and realize the way you’ve been operating just doesn’t work anymore. Now you’re not building forward, you’re scrambling to catch up.
And that’s the real cost. Not just inefficiency, but the fact that your attention gets pulled away from growth and redirected into fixing what should have been built earlier.
There’s a pattern here. Every ISP hits it.
A process works at one level, then quietly becomes the thing holding you back at the next. What worked at one hundred customers doesn’t work at one thousand.
Field operations are a perfect example. When installs are low, scheduling is simple. As volume increases, inefficiencies start to show up. A technician bouncing from one side of town to the other might not seem like a big deal until you realize you’re losing hours every day.
That is real money.
The same applies to billing. Early on, you make exceptions. You adjust dates, handle edge cases manually, and keep things flexible. It works because you’re close to it.
Then you try to automate, and suddenly all those exceptions become problems you have to solve before anything can scale.
Nothing breaks overnight. It just becomes harder and harder to manage until it finally forces your hand.
If there’s one thing that will catch up to you faster than anything else, it’s undocumented processes.
At the beginning, your network map might exist in someone’s head. Your install process lives with your field team. Your troubleshooting approach is based on experience, not documentation.
That works until it doesn’t.
Because eventually, someone leaves. Or takes time off. Or just isn’t available when you need them.
And now you’re trying to piece together critical processes from memory, from conversations, from whatever scraps you can find across the team.
That’s not a position you want to be in.
The reality is simple. If your process cannot survive without a specific person, it is a risk to your business.
There’s always debate around documentation. Do you document everything upfront or wait until the process is proven?
The truth sits somewhere in the middle.
If you document too early, you spend time writing processes that will change anyway. If you wait too long, you risk never documenting at all.
What matters is discipline. You have to come back and capture what works once it’s working. You need a shared place where that information lives. And you need to maintain it.
Because outdated documentation is just as dangerous as having none.
The good news is this part is easier than it’s ever been. There are tools now that make it simple to record workflows, clean them up, and turn them into something usable.
The challenge is not the tools. It’s making the time.
This is where things get interesting.
Do you let sales outpace operations?
There’s a real argument for it. Sales momentum is hard to build and easy to lose. Slowing it down can feel like stepping on your own growth.
But there’s another side. If operations cannot keep up, the customer experience takes the hit. You sell a service you can’t deliver on time, or worse, you overload your network trying to keep up.
Now you’re not just risking new customers. You’re impacting the ones you already have.
So what do you do?
You communicate. You set expectations. You stay engaged with the customer from the moment they sign up to the moment they’re installed.
If there’s a delay, you don’t go silent. You make it part of the experience.
That’s the difference between a frustrated customer and one who feels taken care of.
There’s always a push to remove administrative work from field technicians, and for good reason. Their time is valuable, and it should be focused on installs and repairs.
But the idea that they should have zero admin responsibility is not realistic.
There are things that need to happen in the moment. Notes need to be captured. Photos need to be taken. Information needs to be recorded while it’s fresh.
The goal is not to eliminate admin work. It’s to minimize it and make it efficient.
Because the alternative is losing important details that you can’t recreate later.
What happens if your undocumented processes walk out the door?
It’s not a hypothetical. It happens.
Sometimes it’s turnover. Sometimes it’s something unexpected. Either way, you’re left trying to rebuild something that should have already existed in a repeatable form.
At that point, your options are limited. You piece it together from your team, or you try to recover it from the person who left.
Neither is ideal.
The better option is to never be in that position to begin with.
Operational bottlenecks don’t show up on a network map, but they impact everything.
Growth is not the problem. Growth is the spotlight. It reveals where your processes, your systems, and your planning fall short.
You can either address those gaps early, or you can wait until they slow you down.
Most teams don’t think about these things until they have to. The ones that do are the ones that scale without losing control.
Catch the full conversation on the Bandwidth podcast. Available now on Spotify, Apple Podcasts, and the Bandwidth YouTube Channel.