19. Network-as-a-Service is More of a Financial Model – Tech Field Day Podcast

Welcome to the Tech Field Day podcast, where we bring together a group of IT technical experts to discuss a single idea about key concepts in the enterprise IT industry. This program features members of the Tech Field Day event series experts in their field on a variety of enterprise IT topics. I'd like to take a moment to introduce the guests before we jump into today's episode, starting with Jordan.

Hi everyone. I'm Jordan Martin. I'm a principal architect at Worldwide Technology focusing all things enterprise networking.

Hi everybody. This is Michelin Murphy. Uh, I am a, uh, CSC, uh, consulting systems engineer at Worldwide Technology, also focusing on networking.

I call Jordan when I have a problem, though. My name's Rob Boyd. I'm a professional talker and since we're sticking to WWT, I host a WW t research series among other series of networking content.

Yeah. Glad to be here. And I also call Jordan.

Yeah. Alright, Jordan, what's the answer? First, What is the question?

It depends. Yep, there you go. It depends.

com. In addition to Techron tv, which is another part of the FU Room group. This episode was recorded at networking field A 35.

Let's jump into the premise for this episode. You may have heard the term network as a service recently. It's becoming a very interesting discussion in the networking space, all about how networks can be operated, more like utilities, more like subscription services, but a lot of people aren't convinced yet.

And in fact, a couple of those people are sitting here in this podcast right now because the premise that they've come up with is that network as a service is more of a financial model. Alright, Jordan, you started this. So you cannot, you cannot, you cannot set that stage saying that that's what I believe.

I'm not saying it's what you believe, but I'm saying that you kind of kick started this idea. So can you tell me why you have this idea that network as a service is more of a financial model than an operational model? Because in some situations it is Such as, um, Well let's start here.

Let's start with this, this idea of, of network as a service. I think what people expect is when you talk network as a service, you're thinking like how we consume cloud and compute. Mm-hmm.

I can go pay for the amount of networking that I need, um, be somewhat of a frictionless experience and and be able to just consume networking. And I think that's one end of a spectrum as you talk about it. The other end is we see some major players in the market who want to provide network as a service, but it really is just leasing with a new name.

The idea is I can get equipment, uh, I can maybe pay per user on equipment. Uh, but the reality is I'm still deploying, I'm still responsible for maintenance and operations. I'm still responsible for all of the things except for now I just pay for it a little bit Differently.

So you're saying that if for network as a service, when that comes up, let's say you're working with somebody and you're providing advice, maybe one of your first questions is, well, let's break that down. Yeah. What do you mean?

What do You mean? 'cause you're saying it could be, it could set you up in a bad direction. If you're consulting on a project and you're on one definition and they're on another, that That's exactly right.

That's, I think that's exactly right. And you wanna be able to not just say, what do you mean? But you also wanna know what's the customer trying to solve for.

Mm-hmm. I mean, if it's, it's even much more basic because sometimes you'll have conversations with customers and they're like, I need a, I need a an XI and, and what they're really solving for, maybe they've decided that X is the solution for the problem, but they don't know enough to know whether or not X is really the solution to the problem. So you really, I mean, there's, it's like an onion.

You gotta peel it back and you gotta figure out, I think what, what is the, what's the problem that we need to solve for? I mean no, no. I think in a lot of cases, I think Jordan is correct that, that what we think of as, or what is being branded or marketed as network as a service is a, it's a sub sub some kind of subscription model.

And it's, it's a way to, you know, but Because it still requires physical hardware and connectivity, I mean, the first time the term came up, and I was just thinking through the term just in and of itself, and it feels like it's been around a long time, but it felt a little bit like glomming on to cloud services that were becoming popular and saying, well, how could we be consumed in a more, um, flexible basis. Mm-hmm. Right.

But then I was like, but yet it's one thing. I can host my data somewhere else, but that connectivity is still, like you mentioned the services. So does network as a service imply that you're gonna have people there to run it?

Or that you just want like a leasing model? I just want the equipment on something outta my cap, my opex versus my CapEx. And the answer is, it could be either.

Yeah. Right? Like, and, and, and, and neither one of those is like a wrong answer, right?

Like, to Tanisha's point, like there are organizations and people that are really just looking for a, a, a more favorable financial model, and that's okay. And there are companies who are providing that. Some of the traditional networking companies are providing that as an ability way to consume stuff.

That used to be this huge CapEx expense upfront where I have to pay it all upfront. I have to size it with a five year lifespan in mind, and I'm not paying it based off the number of users. This gives me a little bit more flexible model about how to pay for the same type of equipment.

But it doesn't change the operational model at all. It's strictly a financial model, but we see other entrants in the market. This is where it's new.

This is where I think it gets interesting. Right. 'cause I think this is where Right.

Help make it easier on me. Sure. So, and, and we're seeing it in a couple of enterprise spaces.

So we see it in, in wan Yeah. Specifically like around SD wan. So that idea, we've, we've, we've, you know, kind of flirted with the idea of what's the right way to build a backbone and where do I put my hubs and how do I build an sd-wan?

And the reality is, is the large majority of our traffic now is going to data centers or clouds. We don't have a lot of the peer to peer traffic like we used to have before, right? We, we used to build like these big IP phone networks where I'd have to have lots of traffic to go from branch to branch.

It's kind of going away as people consume that more in WebEx or teams or other places and more cloud delivered models, right? And so now this idea is like, okay, now it's more important about how do I get to the cloud? How do I get to peer organizations, B2B type connectivity?

And then maybe a little bit of getting branch to branch, but it's really branched to data center, branch to cloud. And, and in that we've had different models of doing that. We've built our own connectivity.

We've used things like, uh, fabrics, Equinix, and those types of places where we could do that. And the next iteration of that is, Hey, um, as a provider, I can build that whole backend for you, do the segmentation that SD-WAN does, the reliability, all those additional features that SD-WAN provides, and you only need to have the device that sits at your edge. And now I can consume a WAN that I don't really have to build or configure.

I just have to get to it. And I get to it by sending the traffic to a hardware device. It's just the only thing I'm responsible for.

That's the WAN side. We also see it in the land side. I think it's a little bit harder of a, of an equation, right?

Like if I'm gonna do, you know, wired and wireless access, it's still gonna be equipment that sits on prem. Now you say, is it a shared responsibility model of who manages that? Who deploys it?

I think that's, I think it's a less obvious as a service than the WAN conversation. But I, it's still, I mean, we, there's multiple companies who are trying to solve that, that Access. But it's also perhaps the way to go if you're both a low resource company, perhaps you're smaller, uh, you don't want to invest in the type of talent we have in the room for a tech field day.

Um, and, and, and you're also not sure am I even gonna be in this space long? Or how am I gonna need to scale? Or am I gonna scale here or there?

So maybe I don't want a heavy investment, but yet I need some advanced capabilities that are gonna have to be enabled. By that, I mean, do you consider Meraki like a offering software? It's, it's funny that you bring that up because I think that Meraki was the closest to this.

Yeah, yeah. Right? Because because you have the financial side.

Even, even on the other end of the spectrum, when you talk about the stuff that actually, what I consider real network as a service, like where you're actually getting something different. Yeah. Um, you still have the financial model where I'm, I'm consuming it differently, but the operational model's really different.

It's gonna be cloud managed and centrally managed and opera, which is Meraki's model. Yeah. It's like Meraki, if I don't have to own and deploy all the endpoints, well, 'cause I feel like your network's not gonna be a differentiator to like, depending on what industry I'm in, more than likely I need it to work.

I need it to work well, and I want to be on a base level with my peers. But at the same time, I, my ability to have the top of line IT department doing something unique isn't maybe gonna make, you know, isn't gonna move the, the pile for Me kind of hit on a religious battle, right? Like, like, is a network a differentiator?

Uh, and, and there are people that I consider friends who are very, very smart who say, And you lost some friends. Yeah, yeah. Yeah.

That, that would say absolutely. Like every company, like their network is supposed to be a differentiator. I don't see the world through that lens.

Yeah. I, I think about the large retail organizations of the world, you know, thousands and thousands of branches, their differentiator is the data, not how the data gets around. Yeah.

I need that. I need that to be super consistent and cookie cutter and all those branches. So I can manage it easier.

I can see That being very popular. Yeah. Right.

And so, and so like in these, you know, thousands of branches all the same. Even for that land access model, being able to be cloud delivered, having this real consistent way to deploy it, very, very prescriptive way to way to put it out there, I think is a very attractive option. But That goes back to my original point, which was talking about what problem you're trying that to solve.

That's right. Yeah. You know, for, for the small company that doesn't maybe have the, you know, deep well of knowledge, you know, they're Solving for a talent Shortage, they're, they're, they're solving for an entirely different problem than say the, you know, the, the retail giants of the world who need to have consistent models for every ti every two months when they roll out a new store, that's an entirely different problem.

Yep. But it goes back to like those basic questions that you have to ask, you know, what, what, what problem am I trying to solve? And as soon as you fi, when you figure that out, you can now root around in your, in your, your tool chest and figure out, well, what root around tools Yeah.

That I have that, you know, or solutions that are available to me, um, can help me solve this problem. Or, or do I have to go and innovate something? Plus I want it to be constantly updated.

I want, yeah. I want the latest and greatest when they're at. Although I don't think the innovation changes that much.

Most of our customers we're just trying to get them up to something modern. Yeah. But, uh, No comment.

Yeah. But, but that brings nothing To say. But that brings a good point because historically, this idea of purchasing equipment and having someone else operate, it was entirely the domain of managed service providers, right?

We would hire somebody to come in and rack and stack the equipment and run it and make sure that it's patched and that's, it's not broken. Mm-hmm. But now that model seems to be shifting away.

Yes, there are always gonna be managed service providers, but there are also a large number of people out there that are saying, well, why am I paying somebody else to do that when I can do it instead? And we've seen a large number of, of entrance into the market, whether it's Nile or Meter or Ramen, who are trying to offer that service, whether it's wireless connectivity or, you know, end user connectivity or WAN connectivity. And they're saying, we can do this better, faster, cheaper than the other people.

And you don't have to management or incur the costs. So is this more of a matter of the entrance into the market, trying to capture that existing relationship that maybe they had with a managed service provider and redirect that funding into what they want to do? Maybe, uh, I think that managed service providers good, at least analog or early analog to what this all is.

Because that's, at the end of the day, what we're talking about is mm-hmm. Another organization managing the connectivity or the access for, for your network. Um, I, I think where it shifted is, I feel like managed service provider before was just like, Hey, we need to have this thing.

We're gonna go pay for it. And, and we then we don't need to, we don't need to worry about ourselves. It's something we can do.

And I think that same impetus is there, but the technology has gotten more difficult. And so as technology continues to become more complex, this is what we saw in cloud. Like when you think about what, like a cloud connectivity and how all those services are patched together, that's an incredibly complex environment.

And the reason why it's complex is 'cause you add all of these features and capabilities where you're can just go check a box and make something happen. Where before that would've been weeks worth of work. We want to have that same capability and networking it requires additional complexity.

So I think that what ha what's happening is the scope of the organization that's willing to accept that type of model is growing. It used to be that it was just a very, a smaller organization that just couldn't staff technical, technical people. Now, I think that you see larger organizations who say, our differentiator isn't the way that we, you know, deploy an access portal.

It isn't the way that we choose to do segmentation on our wan. Our differentiator is the way we run our business. And that is the data we collect, how we service the customers, how we do whatever.

Like that's really the differentiator. So if I can pay somebody just like in the cloud, I don't have to become, and I don't have to staff experts in BGP and in overlay networking and all of these other things that as, as things just continue to progress and become more difficult, the idea is I can just essentially outsource it. I can, I can pay someone else to be those experts.

I always wonder about it because when the, who's making those decisions? 'cause if it's a, if it's a company that didn't even understand why they might need one network type versus another perhaps or something, then it gets into, well, if they don't, then they're just saying, well, I just need connectivity. Each individual just needs good connectivity for us to be able to do our thing.

I don't know, I I feel like it's one of those like cloud and maybe it's a follow on to you see cloud grow and then there's a little bit of a correction. Yep. Uh, as people go, wow, that's expensive, right?

Mm-hmm. Um, same type of thing it feels like could happen as network, as a service where you go in because it's easy, you need to get started, but then some reality sets in. But the reality, perhaps the setting in is your maturity is improved so that you're better able to maybe make decisions with or without a network as a service provider to be able to say, we do need to do this on our own, or we don't, we can continue to invest in this pay by the drink.

But at the same time, to that point, one of the things that we see a lot in the enterprise space is this idea. It's not necessarily the sunk cost fallacy, but it's that opportunity cost. If I'm paying on a monthly basis for cloud, what I'm expecting is, is that whatever I want to be available is available.

Yeah. If I wanna turn on 10 extra servers, I know there's a cost to it, but those 10 servers are available for me to purchase. Whereas if I've purchased a million dollars worth of gear from my headquarters, I am stuck at that functional level as long as that gear is in my headquarters.

So a good example of this would be wifi access points. Mm-hmm. If you refresh your access points five years ago, you are two technology revisions behind because you're operating off of uh, uh, wifi 5, 8 0 2 point 11 AC wave two for those of you who are nerds like me.

So we've already gone to AX wifi six and seven, six gigahertz wifi six E and we're on the cusp of seven. And so there are a lot of companies out there that see the bigger numbers and say, we're missing out. Even though realistically speaking, the network's probably fast enough.

There aren't a whole lot of clients that can do seven right now. And I guarantee you nobody has anyone in their desktop or their iPhone. But it's that idea that they're missing out on something they could enable.

You know, this goes back to the days when I was learning how to compile Linux kernels and I was just checking every box because I don't know if I'm ever gonna use this, but I want that in the kernel in case I need it. And I was unfortunately slowing everything down, right? Because I didn't realize that you need to optimize for operational efficiency, which is what you're talking about.

Getting rid of operational inefficiency makes my business stronger. And if one of my operational inefficiencies is the way that my network has operated, we'll get rid of the people who operate it that way. And that becomes a pendulum, right?

Because for a while we push on one end of it and then we come back. Um, that's all of tech. Yeah.

Yeah. That's client Server central versus remote. You know, that's a good idea.

I mean, we've been talking about this a lot this week too. 'cause we're talking about scheduled networks, which is just going back to circuit switch networks versus packet switch networks. I mean, like it's, we we love pendulum.

That's why a different name. Yeah. Right.

But I, I think to your point there, I, I think the challenge that is trying to be addressed in this, it's like the same challenge that we see why like Sassy's taking over and that is like, how do you buy a firewall that can do SSL decryption on a hundred percent of traffic and how do I buy that with a five year horizon so that I don't undersize it today? 'cause this things has to last me until whatever. And you know, I have no idea how many acquisitions or divestitures or things are gonna happen to my company, but I still have to buy a piece of equipment today that's gonna meet my needs for five years.

Or I can pay a premium small premium on that monthly service and scale up and down as I need. Is it really more expensive at the end of the day than putting all of that, uh, CapEx model money out at the beginning to try to make that happen? I think people will see that in the network as well.

'cause we see that same thing with speeds and feeds and capabilities. You're talking about the wireless market. Yeah.

These things. Yep. You can, you can sign up with Apple and get a new one of these every year.

So you're not missing out now you're paying more per month to get that function. But there are people who are happy to do that. But you, you don't own the phone at that point.

You're basically leasing it and giving it back at the end of every year. Whereas someone might wanna buy their phone outright and hold onto it as long as possible. And there's always gonna be both models.

So can network as a service appeal to the people who want to own the network at the end of the day? Is there a model that works for them to be able to have that latest and greatest technology? But if I decide to stop paying the bill this month or this year, or I go with a different provider, you're not gonna come in and rip all my equipment out.

Yeah. Well that was the one thing I, it that disturbed me about some of Meraki was I was on Meraki and my home network because I worked for them, you know, and mm-hmm. They gave me keys.

Well, as soon as I don't have keys, stuff's useless. Yeah. And it's some beautiful hardware, brick.

And it was a brick. Yeah. And, and then that made me uncomfortable.

Of course, when that reality sets in, it was fine when it was, I worked within my free cost basis. Mm. But I always, the companies you're seeing start up in this space, do they, I, I don't know how much you know, but is there, are we seeing a niche that they're exploiting?

Is there, is it, is it mainly the psychology, as you said, of Oh, I've gotta have something that's that's, you know, the best that's meeting my needs on a regular basis? Again, I don't think we can talk about the market as one market because, because we see network as a service kind of going into different spaces. I think that the, the land space is a question mark.

Like the, these companies that are there, they're developing good products. I don't know how they're doing when it comes to, to actually like winning deals and getting into places where there's interest. Um, companies are interested in talking about them, but it's still very, very early.

I think the wan space makes a lot more sense. Not for everybody, but the wan space makes a lot more sense 'cause we're already doing that. The level people who are leveraging, like I said, like Equinix fabric, like I'm gonna go there and I'm gonna use their fabric as a backbone.

I'm already kind of doing network as a service, but I've only like dipped a toe. This just says, let's just go the rest of the way and say, I don't have to build the segmentation to, to them. Well this comment's probably a little off-prem, but the notion of it makes me wonder about what are ISPs doing these days to take advantage of this?

'cause it feels like, it feels like in a loose sense, 'cause I'm not in that market, but that they're fumbling a little bit with regards to how things are shifting to an Equinix and others and the lines are getting blurred. Well, I know, Didn't We have, we can't count on long distance lines anymore. We Didn't.

We have graph on Last year. Yeah. They've done Field Daniel And, and, and they rolled out Tech Field day two years ago.

And you know, when they, they, they came and they talked us to us the, the first year of their rollout. You know, one of, I think one of us asked, you know, where do you see your, your competition? And, and they thought they were gonna have to fight tooth and nail with the ISPs.

And it turned out that the ISPs became, you know, their customers, the biggest customers. Yeah. Because as you know, ISPs grow and they need capacity.

It was easier for them to be able to say, well, we're going to dial into Grapht and use them to help us be, be, you know, more flexible in our capacity. And, uh, so smart. That was, Yeah, the ISP point's really astute.

And I think we can see how it's going to evolve based on what happened around sd-wan. 'cause SD-WAN was gonna eat the ISPs lunch, right? Like they were gonna come in, we're gonna get rid of MPLS, everything's gonna be internet based circuit.

It's good to be like, you know, one 20 or one fifth the cost of what I used to pay for circuits. And the reality is that ISPs came in just embraced SD-WAN and said, you know what? We'll give you SD-WAN suddenly The cost difference.

Yeah, right. We'll, we'll give it to you and it will work great and we can leverage both MPLS and whatever. And like they were able to adjust into that and the ISPs did not lose because of sd-wan.

They're doing just fine. But, But they were do, they were being astute about It. Right.

And I think that network of service will be, will be something similar. And if we, if we really are being genuinely honest when we talk about the wan as, as that, when we talk about what like graph's doing, it's not wildly different than what the old NPLS networks used to be. No.

Where like I go to my ISPI tell them I want a WAN and I have one edge device that connects to them and they do everything in between now. Yep. And they do everything in between.

That is a way oversimplification, right? So if you, if you look at what, like what the likes of grapht are doing in providing connectivity and some of the things that they're doing, they are way more flexible than what the traditional NPLS networks were. And they give you a lot more independence and, and, and tunability.

But it's not a completely separate analog. They're, they're live in similar spaces. But to That point, the reason why an MPLS network operates the way that it does is because of hard rail limitations.

Because when we implemented them, that was the state-of-the-art hardware at the time. And one of the things that we've learned over the years about it, ISPs, is every day they can operate that old gear as a day. They don't have to buy new gear, new gear profitable.

Whereas with a network as a service model, they're paying per month no matter what. So if they can refresh that gear or offer new services that increase their revenue, then as long as what they're paying is less than what they're making, then it's a win for them because they don't have to go out and write a new check every four years for that giant refrigerator sized router that makes them all squeak. And I think that that comes back to the whole discussion that you mentioned about the financial model thing.

Maybe network as a service is just a way of amortizing those big connections that you have to do every three or four or five years. But doing it in such a way as to keep people from squeaking. And so now I can more closely match my outlay of revenue through opex to what I'm getting in profit and it makes the bean counters happy.

So maybe, maybe network as a service is really just a way to make the accountants happier. Aren't we always just trying to work that direction? Yeah.

I think to some degree, yeah. I, again, like I don't, I don't think that any conversation around network as a service can be had without talking about the operational model. And so while I disagree with the premise in, in the saying that it is only an operational model, although you can certainly approach it that way, operational model can be, or I'm sorry, cost model is gonna be part of the whole thing.

Like whether either into the spectrum that you're on, the reason why you go this direction is because you wanna normalize that cost, make it predictable, and, and then you also don't wanna get yourself locked in on hardware that may not meet your needs going forward and you end up with a surprise upgrade or something that you weren't expecting. Well, I think that the premise for this episode that it's just a financial model is of course not entirely correct because network as a service is actually a business decision. It's something that I do to make my business go faster.

Because quite honestly, if we didn't have a business, we wouldn't need a network. But there's more decisions that go into it rather than just operational model or financial model. We have to take everything as it comes and make sure that we're looking at our network as an asset and not a cost center.

And however we have to do that as the best way to transform it. And if you are on the journey or beginning the journey to decide whether a network as a service is right for you, don't forget to pull in the rest of the people in your business to help understand what makes them happy and more capable and ensure that you are including that in your decision making. Because if you don't, I promise you, you're gonna be the one who's being serviced out of the organization before you know it.

That'll just about do it for this episode of the Tech Field Day podcast. I wanna thank everybody for tuning in. com slash podcast.

Until next time, please make sure to share these episodes with your friends. Smart leave a rating and a review on all of our platforms where you listen to us so that people can find us and, and have as much fun as we do recording these episodes. Once again, this podcast is brought to you by the Tech Field Day event series.

com. Until next time, take care.