I decided to start writing about the day-to-day obstacles we face at NEPA Fiber. Before I begin to make regular write ups on our progress, I feel it's necessary to explain the backstory of how it all began and where we are now.
The first time I put some serious thought into starting an Internet Service Provider (ISP) was when I had just returned home after being dropped from Officer Candidates School (OCS) in the Marine Corps. I had endured the most difficult 8 weeks of my life, both physically and mentally, and wasn't sure what was next. Although I didn't graduate from OCS, my time there taught me that I was capable of enduring much more than I ever thought I was.
My parents have been nothing but supportive in everything I do. When I returned home from OCS they encouraged me not to jump into anything too soon, but rather to take some time off to figure out what I wanted to do. I had a pretty substantial tech background; I worked at a datacenter, was a software engineer, and started and sold a small web hosting service prior to embarking on my short-lived journey in the Marine Corps. Like my decision to join the Marine Corps, I wanted to do something meaningful that positively impacted peoples lives. I also wanted the freedom to make my own schedule and not be locked into a 9-5 desk job.
Internet options in Northeastern Pennsylvania have historically been atrocious. It wasn't until most of the country had cable internet that we finally had DSL; and it wasn't until most of the country had access to fiber that we finally had cable internet. I spent my entire childhood using Dial-up(50 Kbps) internet and then DSL (1 Mbps). It wasn't until I moved to Philadelphia that I was hit in the face by what the rest of the country had access to. In 2010 I was likely among some of the very first subscribers to Verizon Fios; sadly that all came to an end when I left Philadelphia and moved back home. It became fairly obvious that although NEPA was once one of the wealthiest areas in the country, we were now living in the dark ages and something needed to change.
After returning home for the second time, I again took notice that nothing had improved. With my experience working in a datacenter and running a web hosting company, I felt I had the knowledge and skill set necessary to be the change I wanted to see happen in Northeastern Pennsylvania.
It's been quite a ride since I started NEPA Fiber in the fall of 2015.
I began by searching for properties in Wilkes-Barre that would serve as our first office. I had two requirements for choosing a location:
- Must have access to fiber.
- Must be located at a high-point that has line-of-sight to much of the city and surrounding areas.
The property needs fiber for obvious reasons and I also needed a location with good line-of-sight to most of Wilkes-barre in order to serve customers using wireless equipment. My long-term goal has always been to deploy fiber, but wireless was a good MVP solution to help kickstart the business as well as validate the business model.
While reviewing Zayo's network map for preexisting fiber, I found out that 65 Public Square, a 14 story high-rise on public square, already had lit fiber on-site. That seemed like my best bet to get started. On November 22, 2015 I sent out a brief inquiry to Jeff, who owns and manages the building;
We're potentially interested in renting some space on the building for antennas. Could you pass on some pricing information?
He replied within the hour. We met the next day and were able to workout a deal that I was able to live with. I was nervous as hell. Between our fiber from Zayo and the building lease I had committed to $2,000/month in overhead with absolutely zero revenue. Everything I had ever read about starting a business told me to start with no less than a years worth of runway; yet here I was starting an ISP, perhaps the most capital intense business on the planet, on nothing more than credit cards and about $2,500 in cash.
Fast forward to January 22, 2016 and Zayo still did not have our service ready for us and money was flying out of the door from equipment purchases and our building lease. I decided to search for a temporary solution to get up and running ASAP. It turned out that a tenant from BurstNet, a datacenter I used to work at, moved in next-door several years earlier. I got in touch with him to buy some bandwidth so that I could get up and running. It wasn't until a month later that we were finally up.
Between when we first went live on February 22, 2016 and when our Zayo link went live on April 28, 2016 we had numerous reliability issues from the neighbor we temporarily bought bandwidth from. They performed a rather half-assed cable run between their building and where our equipment was located; this resulted in the cable breaking on at least 3 separate occasions. In each instance we were down for over an hour. They were instrumental in getting us online early, but I was extremely happy to remove ourselves from their network.
Started From The Bottom
When I started NEPA Fiber I literally had nothing. No equipment, no tools, and no truck. I borrowed my friends truck to transport a ladder or put a foldable ladder in my car when installing our first 7 or 8 customers.
Overtime I slowly acquired all of the installation tools, networking equipment, wireless equipment, and took a loan out to buy the NEPA Fiber mobile. In 2016 alone, I spent upwards of $40,000.
I'm also fortunate enough to have some amazing friends who helped put up our first antennas on our tower in 35 degree weather.
I knew from the beginning that wireless wasn't a long term solution, but it has been particularly frustrating to keep the equipment running issue-free.
It works great when it works, but every once in awhile some of the gear will lock up for seemingly no reason. I'm about 85% sure it's an issue with static-electricity building up in the devices, but there's nothing I can do to resolve it; it's a design flaw.
The other issue with wireless is that we're quickly on a path to capacity and quality of service issues. We have roughly 40 wireless clients all sending and receiving from the same building; that means the spectrum around our primary office is getting fairly crowded and the equipment is becoming more and more susceptible to interference between devices. This is avoidable to an extent by using some smart channel techniques. I have yet to receive an actual quality of service complaint, but I feel that it's only a matter of time before I start seeing some service quality related issues.
One of the necessary evils in bootstrapping a company is taking calculated risks. NEPA Fiber is completely bootstrapped and self-funded. Unlike VC backed startups, I don't have the luxury of trading profitability for growth. It's either make money or die.
One of those calculated risks involved starting an internet company with only one upstream provider. This basically means that when Zayo goes down, we go down. It seems pretty silly for an internet provider to rely on a single upstream provider, but when the cost of adding a second provider is mid-high $x,xxxx/month, it's not really a choice.
My worst nightmare struck. Not once, but twice.
On August 31st, our service had gone down for approximately 11 hours. Zayo had scheduled maintenance on our link for midnight that night, so when I received an alert from our system that our network was down at approximately 12am, just after we received the "Maintenance Start" notification, I had every reason to assume it was due to the maintenance. However, when I woke up at around 6am, our network was still down. It turns out that one of the networking cards on the SONET for our service had failed in the middle of the night. Our service had come up after the maintenance, but then went right back down. By the time Zayo drove up to Wilkes-Barre from their offices in Allentown to service the equipment it was 11am before we were back online.
Exactly one month later my luck struck again. On September 29 at 5:55pm, I received an alert from our system that our network was down. I immediately called Zayo and they told me that they were receiving dozens of similar reports and that they were dispatching a team. At around 9pm that night we were told that construction at the new Comcast building (ironic, right?) had severed conduit containing a 144 strand fiber cable and that repair was in progress. Our service was finally restored on September 30 at 9:45am. A total downtime of approximately 16 hours.
Needless to say, our customers were not happy. I did my best to be transparent and update our subscribers on the progress being made to restore service. In the end, we somehow didn't lose a single customer; although two did threaten to cancel.
When we first launched, adding a second upstream fiber provider was out of the question. It would have made it impossible to turn a profit. By the time these incidents occurred we were turning a small profit and I decided it would be worth the risk to sign a contract for a second provider. On October 3, 2016, I signed a contract with Level 3 for a second 10Gbps uplink. It was a huge risk, but I'm pretty happy that I made the decision back then to do so; the build got delayed and we are expected to receive delivery of the second uplink towards the end of April, 2017.
NEPA Fiber is finally coming to the point where running physical fiber in the ground is not far off. Our revenue is strong and the community has taken notice to our superior service.
We've signed contracts with many of the high-rise buildings in Wilkes-Barre to bring fiber to their tenants. This is a turning point both for NEPA Fiber and for business in the area. Dozens of companies have contacted us to bring fiber to them and many of them are actually moving their office space to Wilkes-Barre City just to obtain our service.
Our biggest obstacles in the near future will be getting approval from the city to run conduit/fiber in the streets and capital. We have the credit and capital necessary to perform construction of our first mile; the question is where do we go from there? Investment would be nice, but I have had zero luck with that to-date. The other option is to grow slower than I would like, but that would provide our competitors with a chance to catch up to us.
We also recently acquired some amazing space in "The Bank" for our datacenter. There's quite a bit of work left before we're ready to move in, but I'm excited to see where that takes us.
Where We Stand
Today, we service approximately 50 customers in or around Wilkes-Barre. By mid-2017, many of our in-flight contracts will begin to kick in and we should double or triple that number. 2017 is definitely the year that we're going to change the game. From there, the sky is the limit.
My next goal is to open another PoP (Point of Presence) in the city of Scranton. Scranton has a much larger concentration of businesses than Wilkes-Barre, but we would also be facing off against Comcast.
Today NEPA Fiber offers wireless and fiber internet services to business and residential customers in Wilkes-Barre, Kingston, and Edwardsville.