Jeff: Hello there, and welcome to the All Day Digital Podcast, where we talk to industry executives and thought leaders to get their perspective on a wide range of factors shaping the communications industry. This podcast is brought to you by CoBank's Knowledge Exchange group, and I am your host, Jeff Johnston.
On today's episode, we get to hear from Marcellus Nixon, vice president, internet services for United Cooperative Services. Marcellus has a unique perspective on the factors electric distributors need to be aware of when getting into the broadband business. He gained this perspective by spending the vast majority of his career in the telco industry. Now that he is working for an electric distributor, he is seeing firsthand how these companies can get tripped up. Electric co-ops are playing an increasingly important role in bridging the digital divide, but the old adage of, you don't know what you don't know, has some relevance here. Without any further ado, pitter-patter, let's hear what Marcellus has to say.
Marcellus, thank you so much for joining us, and welcome to the podcast. How are you doing?
Marcellus: I'm great, Jeff. Thank you. Thank you for having me.
Jeff: Oh, listen, absolute pleasure to have you with us today. I'd like to get your thoughts on the role that consultants play in terms of helping electric distributors get up to speed on thinking about broadband, and designing the network and so on, and are there other things that these folks should be thinking of post the time they spend with the consultants. This could be business management, this could be network operations, stuff that maybe aren't appreciated or understood, because from your perspective, you're coming through a telecom lens, and I think you've got a unique perspective on this.
Marcellus: Yes. I'd love to talk about that. I think the larger consultants are really good at creating the feasibility studies. That's basically, will this work? How long will it take? What is the percentage of your customers or members that will be covered in this project? Then how much is it going to cost? I think they do a really good job of that.
The one thing that I would recommend is, anybody going through a project like this, is that, or bandwidth consumption by the consumer, that they use the FCC minimum of 25 megabits download and 3 megabits upload. Although some people may not need that and some people need a little bit more, I think that should be your gauge, because that has a direct impact on your capital spend and your operating expense.
I think, getting to the latter part of your question, I think the biggest concern is, it's one thing to build a network infrastructure, but something else in totality to run it on a daily basis. You have things like, as you continue to bring on new subscribers as traffic patterns change from your ISPs, what is the member experience like? Let's just say, for example, on our network in particular, 37% of our traffic that our members are demanding, is Netflix traffic. If you're not prepared for that traffic, it could really create a situation. That's 37% of your customer base or member base that's going to be negatively impacted if you don't think about that in the right way.
What we do here is really look at the flows of the traffic on our network to determine that, to see and if we're looking at the flows, we're looking at how the members are using the infrastructure and the service.
Jeff: That's interesting. I'm not surprised to hear Netflix is such a huge driver of bandwidth. How do you manage that? What are some of the things that, so you need to put in place to anticipate maybe these trends, and then what do you need to do from a network architecture standpoint to be able to support these different spikes in traffic?
Marcellus: That's a great question. Initially, we had a single router and two head ends, which are our internet access points. Then we have quickly evolved the infrastructure. Once we could start seeing where the traffic was destined to and coming from, we started to move to our two best industry practice, internet Edge platform, meaning we put routers or router infrastructure in the telco hotel building, so in our case, it would be Dallas. Dallas has two telco hotel buildings, and the two are the biggest buildings of its like in the world. Once we put our infrastructure in those locations, it allows us to connect to the networks where we see our members' traffic being served. Those would include Amazon, Google, largest content delivery network in the world called Akamai, and of course, Netflix. Facebook, Microsoft, the typical players that you see, and then a lot of the gaming networks. Our network is really popular with gamers, and it's because we can connect directly to those networks instead of sending that traffic through one of our ISPs to reach.
Jeff: Got it. Let's just go back to the consultant dynamics. As you mentioned earlier, they're really good at the feasibility study and analysis, helping EDs get that one-foot forward in this process, but then how should they think about, from a staffing perspective or an expertise perspective, or maybe an outsourcing perspective, how should they think about post feasibility post that consultant interaction to when we really start getting going here where we're tearing up ground, we're stringing fiber along poles and so forth? How should they think about their organization and strategically and operationally to make sure that they're in the best position to succeed, besides hiring smart people like you who have a lot of telecom, and maybe that's the answer.
Marcellus: Thank you. I do think that hiring people like me can definitely help. I would say that just at a broad stroke, you have to over-index on the support mechanism for this infrastructure. What I mean by that is, a lot of them will look at you need X amount of support people per percentage of members. You need to ask questions like, do you want 24 by 7 support? Members can call in 24 hours a day, 7 days a week. Whether you're going to be supporting in a large population of commercial members. Commercial members cannot be handled in the same way as residential.
No disrespect to someone's household, but when you interrupt traffic to a commercial entity, that's their livelihood. If they're down, if they have an impairment, they need to be serviced right away. You need to have a more highly skilled support mechanism to support those types of members. These are things that need to be thought about, but I would say at a broad stroke, you can never go wrong over-indexing on support.
For us, for example, we know we need at least 10 people to be able to answer 24 by 7 calls from members during hours of between 6:00 AM to 6:00 PM. Then we can pull it back a little bit in the evening. We need to have a network operations center that has at least two people every shift. Every eight-hour shift, they need to have two people. That doesn't mean that they're sitting there, Jeff, just monitoring and looking at things, they're actually doing things like software upgrades on different pieces of infrastructure, as well as looking for impairments to service, or answering calls from commercial members.
Jeff: Got it. No, that's really, really helpful. Let's move the conversation a little bit further along here, and I'd love to get your thoughts on, and maybe this is, I guess, a little bit of a part of the feasibility analysis, but I'd certainly like to get your thoughts on, how should EDs think about what markets they should enter? How should they think about competitive overbuilds? These folks aren't used to competing in some cases highly dynamic competitive markets. What are your thoughts on that? What have you learned? What have you observed on how lessons learned that might be able to help some of these folks out?
Marcellus: I would say the one thing that you don't want to do is, you should take a stance on, you only want to connect to members who are either unserved or woefully underserved. What I mean by that is, if they have a choice of another provider, and the other provider can offer speeds up to one gig on fiber, that's probably not a situation you want to go into and compete head to head.
Just to give you an example on cost, for us to go into, let's just say a hundred home subdivision, it's about a $400,000 cap ex hit for us. What you don't want to do is go into a situation like that and only get maybe your take rate is, instead of-- it has to be for us, it has to be 32% of the overall people in the subdivision. We only get 3%. That means you have this stranded asset that you deployed all this infrastructure, but you're not receiving any return on that invest. I would look for areas where in your territory, where there is no coverage at all from a fiber provider, and prioritize those, and then see how it goes in the other areas, if you go into them at all.
Jeff: For example, if there was a market that had DSL coverage that was neglected by the telco, maybe that could be an opportunity?
Marcellus: Yes, that will be a great opportunity. If there was one that already had a fiber provider, and there's fiber to the home already, or the fiber to the business, probably not a good situation for you.
Jeff: That makes sense. Then just maybe one more question on the technology side, fixed wireless versus fiber. Obviously, fiber is the gold standard. We want to do fiber, it's future-proof. Are you coming across or do you hear about instances where if there's an underserved market, does it make sense to do fixed wireless in your case, or do you think you really should push fiber as deep as you possibly can into the network and stay away from fixed wireless? Any thoughts on that?
Marcellus: I would say as a general rule, you're right, most people want fiber, it's a better strategic trajectory than fixed wireless, but sometimes it just doesn't make economic sense for you to run fiber. If you have a location, for example, like in the western part of our territory, where homes could be miles and miles apart, you would break the bank trying to reach some of these homes, because they're miles off the main road, and they have vast areas of coverage, even within their property. Then you can get to the next home that will be in the same situation.
Whereas you can erect a fixed wireless asset, have a line of sight to a fixed wireless monopole and you can reach those homes, and they can reach speeds up to 100 Megs. It's very important to look at your terrain and your customer geography as a whole, and then you can decide to always look at what the economics are. Let's just say you have a situation, for example, where it cost you $4,200 per home passed for fiber, that's an area you probably want to start looking at fixed wireless. There are some philosophies from some other top consultants, and I can't blame them for this. They don't like fixed wireless at all.
At the end of the day, you have to do what's economically feasible for your organization, in your particular business case.
Jeff: I know that the knock against fixed wireless, when we just had using unlicensed Wi-Fi spectrum and 5 gigahertz with just basic Wi-Fi gear was no throughput speed and wasn't great, modulation wasn't great, capacity really wasn't there. Does the introduction of CBRS spectrum and using LTE technology, does that, in a significant way, strengthen the argument to do fixed wireless, or maybe not so much?
Marcellus: No, not for what you would see in the whole broadband space, in the whole broadband use case. That's more of a carrier type of technology at this point. That could change, but not what we're using for our fixed wireless deployments.
Jeff: Great. Then when we think about EDs deploying internet service to their members, so obviously one revenue stream that they could enjoy, but are there other things that the EDs could do with that network to provide alternative sources of revenue that might make the business case look a little bit better?
Marcellus: Yes. Some of the main ones, and this is not an input to our particular business case, and I don't think feasibility studies should even include this, but you should always look toward that high, being able to provide, let's just say, for example, a big long-haul provider might say, hey, you guys have this fiber plan going from this point to this point. Can I lease two fibers or four fibers or eight fibers from you guy? It's a 20-year IRU, meaning that I'll have it, it's an asset on their books for 20 years, and they basically pay you either a lump sum fee or a monthly recurring. That could help you drive fiber deeper into an area where you were going to put fixed wireless, as an example.
Another one would be you have fixed wireless monopoles, and you might have a mobile provider that will come to you and say, "Hey, this is a dark area for us, or a spot where we have a lot of complaints about lost coverage. Could we mount our radios on your towers in this particular area to cover that?" Again, same type of scenario. It's a 10-year, it's a 15-year, it's a 20-year IRU, and it's a revenue stream that did not come from your members. It will help you to bolster your infrastructure and provide a better service for your members.
Jeff: I guess you could also maybe even throw, it's on the come I guess, but you hear a lot about edge computing. That could also be a potential future opportunity, but maybe not one you want to factor in, as you mentioned, to your business.
Marcellus: Yes. If you look at how most electric co-ops are building their fiber infrastructure, they're putting their termination devices at their electrical substations. They take it from the general utility, and then they fan out and distribute it out to homes and their members. One of the things that we think about a lot is, you can't get better performance in putting all of your edge compute power at the edge of your network, whether that would be beginning at the head end location for us, which will be here in Burleson, Texas or in Stephenville, and then eventually started working around where that has compute power is placed in that specific local market that only services one particular small area. I think that's the way that most co-ops look at their infrastructure going forward. It's always about moving everything out to the EDs.
Jeff: It sounds like you're a believer, I guess, in edge computing, you think that that really could be an architecture that really starts to take off down the road?
Marcellus: Yes, and if you look at most telco-grade networks today, even the top tier ISPs, they've long adopted a methodology of moving much of their S compute horsepower to the edge of the network, or the customers that are actually accessing.
Jeff: I want to just change things up a little bit and talk about culture, and get your thoughts here, because it's long understood that electric co-ops are hypersensitive, hyper-focused on making sure that their members have the best service, 100% uptime, that's really a big focus of their culture, from what I understand. Cable operators, if you look at some of the customer satisfaction ratings, they don't always rank as high as maybe they should, certainly not as high as electric co-ops.
Maybe just talk a little bit about, as this has been deployed, how do you make sure you keep that same kind of customer-centric culture in the telecom part of the business that exists in the electric co-op side of the business?
Marcellus: Yes, I think, first of all, I look at how we structured our team in particular. We brought electrical culture over to the internet side. If you look at the first managers that we brought over to build the internet services team, they came from the electrical side, so they bring that culture over. You can go out and hire people off the street, but you don't know if there'll be a fit to your point with your culture, and they may not service your members in the way that they-- on which they're accustomed.
I think that's key, bring some of the electrical, or utility infrastructure or customer service over to the internet site and staff it that way, and then the other thing is, I think you have to really-- again, over-index on support. That has to be your main thought if you have to resource properly from a support perspective, or with workflows that make it easier for your members to do business with you or educate them, have meetings where you're talking specifically to small demographics of people, or even having, in contrast to the big telco, having a person answer the phone for them that has the same dialect that they have. I think that's extremely important.
Someone might tell you that I live on County Road, blah, blah, blah, and you remember last summer when the lightning hit the tree at the corner, make that left, and if you have a person that's from that area, they're like, "Yes, I know where that is." That means a lot to that person on the other end. Those are the things that we try to do, and we try to bring that culture over.
Jeff: Yes, I would imagine that the market penetration rate for most EDs is probably pretty high in this market because of those reasons, the reputation that they have, the customer service, and so forth.
Marcellus: Yes, it is. You always have to be looking at, for example, one of the things that our CEO is very sensitive about is, how do we reach our members? Some numbers will read a monthly magazine. They go to the mailbox, they get it out, they read it from cover to cover. Some other members have magazines stacked up, and they will never read them. How do you get through to them? You get through to them through email, through text messages, through our Facebook posting, whatever. We're constantly looking at ways to reach our members, regardless of what media we need to use.
Jeff: That's great. That's wonderful. Well, we're almost out of time here, Marcellus, but before we end it, is there anything that we didn't cover today that-- any points you think that should be made that we didn't cover that you'd want to share?
Marcellus: I think we covered the main points, but one last thing that I would like to say is, I just think that the consultants out there are very helpful, but I do think that you should go into this with your eyes open if you're looking at building an internet infrastructure to provide service to underserved or unserved areas. Keep in mind that, at end of the day, it's not the consultant, it's you working alongside the consultant, because without that, you'll have what the consultants can do, and there's a gap in which your members need to have a smooth serviceable workflow, and that's the most important.
Jeff: No, that makes sense. That's great advice. Well, again, thank you so much much for being a part of the podcast today. This was incredibly an enlightening. I and the CoBank team really appreciate your time. Thanks for making it.
Marcellus: My pleasure. Thank you for having me.
Jeff: A special thanks goes out to Marcellus for taking time out of his day to share some great insight. I think the biggest takeaway from my conversation with Marcellus is to make sure electric co-ops really focus on how they're going to architect their network to optimize speed and performance for their customers, and to also make sure they over index their network in customer support operations.
Things like workflow management, redundant connectivity, data center integration, are very important factors to consider and plan for.
Hey, thanks for joining us today, and please watch out for the next episode.
