What Matters Most at Every Stage of ISP Growth

Show notes

Every ISP is trying to do several things at once: grow, improve the network, keep customers happy, and stay efficient. What matters most depends on where the business is. The priorities that fit an operator with 500 subscribers are not the ones that fit the same company at 5,000 or 50,000.

On this episode of Bandwidth, Georgette Lopez-Aguado, Rick Seemann, and Larry Weidig walk through three stages of ISP growth: the initial infrastructure buildout, network densification and service expansion, and growth through new markets or acquisition. At each stage they cover what to prioritize, what breaks, and what gets too much attention.

Stage One: You Are Building a Business, Not Just a Network

Asked for the single most critical objective during the buildout, Larry picked physical deployment speed. You have to get the network up. Rick took it as a mindset: progress over perfection. Early on you rarely have the luxury of buying the best of everything or engineering for infinite scale at every turn, and an operator who waits for perfect never launches. Taken too far, that mindset creates cleanup debt, so it needs limits.

Georgette named what she sees as the biggest mistake at this stage: believing you are just building a network. Passing homes and adding towers matter, but customers do not buy the fiber. They buy a reliable experience. That is why she builds for repeatability from the start, with standardized installs, clean customer records, documented workflows, consistent billing, and repeatable customer communications. Those habits are very hard to retrofit later.

Rick agreed, with one caveat. A first-time operator will figure some of it out as they go. The discipline is that once something works, you turn it into a repeatable process.

The Early Mistakes That Get Expensive Later

Larry listed two: overengineering, which stops you in your tracks, and poor tracking. Georgette added that operators underestimate operational complexity. "We'll clean up the data later" and "we'll automate when we're bigger" both tend to cost more than doing it right the first time. She also sees operators buy too many disconnected tools too early instead of choosing platforms that can grow with them.

Rick went the other direction: thinking too small. Things can scale faster than a business plan assumes. When a more open-ended option does not cost much more time or money, leave that door open.

What Gets Too Much Attention Early

Rick's answer was gear selection. It matters, but putting all your attention there means neglecting operational processes, and a perfect vendor choice only gets you so far. Over-prioritizing any one area at the expense of the rest is the thing operators regret.

Larry's answer was automating or building custom solutions too early. Automate once you know exactly what the process is and it has become repetitive enough to be worth it. He also warned against marketing to an area you cannot serve yet. In his own experience a utility refused to bring more power to a tower site after customers had already been told service was weeks away, and the team had to find a different location.

Georgette's answer was subscriber count. If installs take weeks, bills are wrong, and support cannot answer basic questions, faster growth just accelerates operational debt. In her words, healthy growth beats fast growth when the experience suffers, because you get one chance to establish the relationship.

What a Strong Foundation Looks Like

Larry kept it practical: clean data, accurate tracking of inventory, serial numbers, and MAC addresses from day one, and installation standards that are up to par quickly, because the install is one of the few times you are in front of the customer.

Rick chose flexibility. Standards-based equipment avoids locking you into one vendor, and shorter contracts the first time around let you confirm the fit. Planning matters, but if you also build in options, a wrong plan does not sink you.

Georgette described a three-legged stool of people, processes, and platform. Hire people who embrace processes rather than heroics, document while the team is still small, and choose systems at 500 subscribers that can still work at 50,000. Larry added a reality check: some platforms built for that scale are not priced for a 500-subscriber operator, so you may need to plan to grow into them.

Stage Two: More Customers Without More Chaos

As the network densifies and services expand, Rick put operating margins at the top. If it costs more and more to keep the lights on as you scale, growth makes you less profitable, and that trend only ends one way. Georgette framed the same problem from the operations side: the challenge is no longer adding customers, it is adding customers without adding chaos. With more services, larger support teams, contractors, inventory, and payment operations, complexity grows faster than subscriber count.

Larry noted this is the stage to focus on ARPU, optimize operations, and bring in things like automated provisioning, since the processes should be settled by now.

What Starts to Break

Manual provisioning goes first. Ten field techs trying to open the same spreadsheet to enter a MAC address does not scale. Rick pointed to the one-off patches made along the way, like an undocumented firewall rule on one tower or billing exceptions that lived on one person's sticky notes. When you try to standardize and automate, you find out how many edge cases have quietly accumulated.

Georgette said communication breaks. Departments start optimizing for themselves instead of the customer. Engineering says one thing, support says another, and sales promises something operations was not ready for. Customers do not care whose fault it is. They see one company.

From Building to Optimizing

The focus moves from expansion to penetration. Larry said this is where scheduled, preventative maintenance has to join break-fix work. It adds cost, and it pays for itself over time. Rick described the capacity side: as passings turn into customers, links fill up and traffic concentration becomes a constant chase, which makes bandwidth and routing optimization a skill you now have to practice. For Georgette the question changes from "can we install?" to "can we install profitably?" and every repetitive manual process should be questioned.

The pressure at this stage shows up as capital, margins, competition, support volume, and investor expectations. Leaders have to stop working in the business and start working on it, and data becomes strategic: why customers churn, which neighborhoods create the most tickets, which products create the highest lifetime value. Larry flagged support staff burnout when support does not grow as fast as the customer base, and working capital strain when buildout costs run ahead of recurring revenue.

Stage Three: New Markets and Acquisitions

Rick opened this stage by saying not much transfers automatically. His word for it was compatibility. A service catalog, a culture, and a playbook that worked in one market all need to be reassessed in the next. In an acquisition, culture belongs in due diligence alongside the books and the network, so you know what you will need to invest in before you sign.

Larry focused on system consolidation and carrying over a strong culture and identity. Georgette called alignment the biggest challenge. At this stage the hard part is people, not technology, and if leadership does not align everyone around one operating model, the complexity multiplies.

What gets harder to control? Larry and Rick both said data integrity: without a clear record of truth, even everyday workflows break down, especially when two systems run in parallel. Georgette said consistency in onboarding, support quality, billing, communications, and brand. As Larry pointed out, those are the same answer applied to different parts of the business.

Priorities then conflict between centralized efficiency and regional autonomy, and over which market gets capital and upgrades. Every market believes it is a little unique. Leadership has to decide where flexibility is allowed and where consistency is enforced.

What Separates ISPs That Scale Well

Rick's honest answer was experience. No two acquisitions are alike, and those who have been through it before get more predictable outcomes. Georgette's was intentionality: successful acquirers build an integration playbook that defines what gets standardized immediately, what can wait, who owns each workstream, when customers hear from you, and what success looks like 90 days later. She also sees AI helping leaders spot patterns across markets, such as emerging churn risk, network trends, and support bottlenecks, once the business is too large to know every customer personally.

Larry's advice was to reach a consistent state as quickly as you can. Do not run parallel stacks longer than necessary, align the networks, and unify the BSS/OSS into a single system, but do it in a controlled way. Expecting everything to migrate by the next billing cycle is fooling yourself. Rick added a caution: a playbook that worked once should not be applied blindly, because diligence will show which parts fit the next deal and which need rethinking.

Georgette closed with the thread that ran through all three stages. Each one needs a different version of leadership. What got you from 500 subscribers to 5,000 will not necessarily get you to 50,000, and the people, processes, and priorities have to evolve along with the network.

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.