Analyst Roundtable
Sign up for Daily CommTech Brief here.
Session Abstract:
Its time for our analyst roundtable at the end of day one of the mobile event in Barcelona. Joining this roundtable to offer the innovations and challenges in the industry are, Matt Hatton, Founding Partner at Transforma Insights, Roy Chua, Founder of AvidThink and Patrick Kelly for Appledore Research.
Executive Speakers:
Matt Hatton - Founding Partner, Transforma Insights
Roy Chua - Founder and Principal, AvidThink
Patrick Kelly - Founder & Principal Analyst, Appledore Research
Transcription
ANALYST ROUNDTABLE
Abe Nejad: It's time for our analyst roundtable at the end of day one of this mobile event right here in Barcelona. Joining this roundtable to offer the innovations and challenges in the industry are Matt Hatton, he's founding partner at Transforma Insights. Next to him is Roy Chua, he is found founder of AvidThink. And next to me is Patrick Kelly, for Appledore Research.
Gentlemen welcome.
Thank you.
Abe Nejad: Well, thanks for being here. At the end of day one right here at the event in Barcelona. I appreciate everybody's time. You guys have had a chance to sort of roam the floor and see the state of play of different sectors of the industry, if you will. Roy, I'm going to start with you, if you don't mind, cloud-native, everything open as we've seen in recent years, and what have you seen that you think we would be surprised by?
Roy: Yeah, so I think, what's been great is as you say, cloud-native and open everything, open transport, open ran, open and disaggregated. And we've heard that before in some aspects and cloud native everything, but I think what's different for me is when I go around and have meetings or chat with the folks of the booth, is that it's not just talking about it anymore, but they can actually have concrete discussions with me around the challenges that they're seeing in going cloud native or the challenges in running an open architecture.
And so I think that's the difference between just saying it and marketing it and actually doing it. And I think I'm actually quite heartened by the fact that we are seeing multiple vendors and the carriers talk about the real-world challenges that they're trying to overcome in both, going cloud-native and trying to open everything up. So, absolutely.
Abe Nejad: Matt, partnerships particularly with Telco’s and hyperscalers, what's the state of play there and what do you expect to see in the near future?
Matt: I think the big question really is how much of the telecoms industry did the hyperscalers want to eat. You mentioned about cloud-native and you mentioned about Open RAN and those cloud hyperscalers who've got a lot of interest clearly in this. And this is an opportunity to kind of own a bit of that space, but everybody's still going to want to try and harness them. It's kind of the beast that they think they can ride, but it's still always going to be a challenge because they've just got the scale. I mean, the clues in the name, right. They've got the scale and they can sell globally, whereas telcos by definition have a more narrowly defined space that they can sell into. So it's going to be challenging for the Telcos to really well ride that tiger, should we say.
Abe Nejad: Yeah, Patrick anything you think we'd be surprised by that you noticed?
Patrick: Our focus is on network automation. So that's where we put our focus. A lot of activity in the AI and ML space. It's not just science projects. We're actually seeing some movement and investment in that area around capacity utilization, spectrum utilization, things like power management and automation of some of the activities in the operators environment. So quite a bit there. I think it goes in line with what Roy and Matt are talking about. So in the open ran, you're seeing sort of this aggregation actually occur out into the ran now instead of the data center. And so you're seeing a more rich ecosystem. There are more companies sort of coming into the space in addition to the incumbents. And that's opening things up, giving operators some options on who they're going to source from in the market. So those are the things that I think we see at this show.
Matt: I just want to say, I think that means it's exciting times. We've got so much diversity and so much stuff happening. One of the big things, one of the big messages, one of the big themes is going to be about private networks. That's just another option people deploy in their own network or have it deployed for them or whatever. And you've got so many diverse vendors getting involved in the space in such a number of different ways that you can architect the networks and provide the services. I don't think there's been a more exciting time to be in telecoms that I can remember. It's fascinating times, and we don't quite know how it's going to settle down, even us analysts. We'll tell you when we know, but we're still watching to see how things pan out.
Abe Nejad: Right. Patrick, you mentioned automation, that's sort of your core competency. Where are we on the timeline, if you will, as far as automating the network, automating processes?
Patrick: There's aspects that have already been automated. So if you look at SON and the RAN, and even Roy just dis a lot of work in the SDN, we've seen a lot more automation occur there in those technology domains. I think the tougher bit is how do you look across a multi-domain, multi vendor service orchestration environment, and start to automate there. And I think that's really where the industry is still maturing if you will. But I think there's been progress. And you're seeing that both in the orchestration layer. And as I mentioned, you know the term observability gets used, but it's taking advantage of all the data that's in the network, trying to normalize that and then act on that data, both in what we would call assurance or fulfillment based functions that is how operators thought about that workflow process in the past.
Roy: Yeah, agreed. I was going to say, yeah, I completely agree with Patrick. And we're seeing that automation going from domain to multi domain across the board. And I think one of the things around automation assurance as Patrick brought up is when you look at someone like Dish or Rakuten for instance. If you ask them what's special and what they're doing is different, very, very often we look at assurance last and in this particular case, like in Dish, they look at assurance first. And so you're building the test and the automation first and then fitting the services in there after the fact.
I think that's quite innovative. And speaking of Cloud-native just talked to [06:18 inaudible] earlier today, and he was talking about using the technologies and the practices from the hyperscale and the web properties like KIOS engineering, or parameters fuzzing to test telco systems, which we never thought about doing that before, but he's now working with the Kubernetes layer and changing the premises on top. Automated, completely automated as part of a telco delivery platform. And I think that is exciting from my perspective.
Matt: Can I just jump in on that one as well? Because there's an interesting analogy with the IOT space. The, IOT is what I cover and it is a fascinating space. And this attempt to imitate what the hyperscalers have done absolutely permeates through that, or a need to imitate what the hyperscalers do, because you can't sell a device that's only going to generate a dollar per year by sending a sales guy out to meet somebody. That doesn't make any sense. So it's got to be automated in turn of the onboarding of the devices, the management of the devices, all of the processes associated with it have to have that automation.
We talk a lot about hyperscale, IoT connectivity providers, imitating the hyperscale approach and really being able to scale up to hopefully tens, hundreds of billions of connected devices, but who knows where it's going to end, but certainly bigger scale than they can deal with at the moment.
Roy: Agreed. It's the mindset changed. I mean, Twilio should never happened if the carriers had done it right in the first place.
Matt: Absolutely. I mean, Twilio's a fantastic example. They start from the other end, okay, what's the services going to be delivered and then how do I work back from there rather than I got network? How can I fill it?
Roy: Exactly. That's exactly right. And that cloud-native thinking, it's not just cloud-native thinking. I think it's just that sort of the web and the enterprise thinking is starting to permeate the carriers. It's starting to think like that. I think that's a positive sign in general.
Patrick: One thing that I would just add even beyond the show, is in order to sort of make that transition, we're seeing business models change a little bit, so where you would essentially buy from a commercial supplier and deploy it on premise. It's now moving to more of a software as a service or telco as a service, or even something like testing as a service, you can start to deploy in that method. And then make the transition to support these new services and these underlying technologies.
Roy: Yeah, that's right. The ASS movement is a real one.
Matt: But is that also partly because people prefer buying on an OPEX basis rather than buying on a CapEx basis, is there an economic incentive or a financial incentive to it, rather than it just being that's the better way?
Patrick: Ironically, the procurement model hasn't changed a whole lot.
Roy: They want to say you give me a CapEx type price so I can manage it on my own because in some telcos, financial engineering is looked upon as a strategic option for them, right. It's like, oh I have great financial engineering and so I can look good. So let me buy it and I'll deal with it. As opposed to tending the OPEX, they want to see CapEx because CapEx in some cases makes it look good.
Patrick: I think the bigger driver Matt is more like you can get rid of the integration tax by going as a service. And then you start to ramp up from there. At the end of the day, I mean telcos need to look at their profitability, because they're not doing too well against hyperscalers on a profitability metric.
Matt: No they're not. But the question is that a fair analogy? If you are in a hyper-regulated triopoly situation, is it fair to compare yourself with hyperscalers who, let's face it there's a lot of survivorship bias to say, well, AWS and Microsoft done very well. There's a whole bunch of unicorns who fail to really take off and telcos maybe just have to accept that it's more of a steady as she goes kind of a business rather than it's going to see that accelerating.
Patrick: I wouldn't want to make that argument to the shareholders.
Matt: Well it depends who your shareholders are. What you might see is activist shareholders in a weird way getting involved in saying, you know what, this is a utility business. We're happy making 5%. So adjust your operations accordingly. And we are happy taking that rather than just making all these bets on different technologies and trying to be something that you are not.
Roy: It may be possible but I think egos will not allow it. We know that. Agreed. But I think the digital [11:05 inaudible] of the world, the innovation, or even SK telecom. I mean, SK telecom is clearly innovative. So I think with the new technologies coming in with a more open ecosystem, and as you say, hyperscaler technology in the platforms that you could leverage and maybe if you do them the right way, you can improve your margins.
Patrick: I think the other counter argument is if you look at what Rakuten's done, particularly with Symphony, I mean their metrics are phenomenal, in terms of turn up and supporting those service on a long term basis. And you look at that from a cost basis, point of view. They're scaling, but they're also delivering on the financial side with some of those metrics.
Matt: To a certain extent though. It's a reflection of being able to, of not having to support a whole load of legacy systems. If you start from greenfield, it's all easy to do. Reliance Jio did it in India. Now I worked at Three UK. Three UK was a Greenfield 3G operator. They didn't do it because 3G let's face it there weren't the [12:12 inauedible]. It was basically garbage as a technology, but you've seen with 4G and effectively Rakuten, with 5G kind of. And so I think there's a benefit associated with not having all those legacy systems. You've got to pick your way through it, if you do it as a greenfield operation.
Matt: Yeah. Fair point
Roy: That's fair. And then now the question is how do you engineer it so that you have the legacy co and NewCo? But that's a different thing.
Matt: But that's what every company has. [12:49 inaudible]. How' you disrupt the market at the same time as running an existing operation.
Roy: There is no easy answer.
Abe Nejad: I wanted to touch on two terms that you're seeing on the floor that we're all seeing on the floor, the ran intelligent controller or Ric or the Ric something we should talk about and touch on and also the metaverse or metaverse networks or metaverse ready networks, however you want to term it or whatever terminology you want to use to apply to it. Let's start with the Ric. SDN certainly has been around for quite some time. What was the impetus for this acronym, Ric or the ran intelligent controller? Where did that come from and why, or why are we talking about it now because it's been around?
Roy: It has been around. So you are correct. The whole software defined networking movement, as we've talked in the past about and the desire to do programmatic control of the underlying network. I think when there ran started the segregation, started being open. I think there was a realization that there was a strategic control point, again, one more time within the ran that could add value. And that's how the Ric came into being. The ran intelligent controller, the ability to run different algorithms on the ran and obviously the Ric is broken into two parts. The near real time and the non real time, not the best naming, but I didn't name it. So it's very confusing. They're both NRT. But the goal of those things is to open up the ecosystem to basically have the ran as a platform and to allow the ecosystem to be opened up.
And so that application developers can come in and add value without having to build a Ran platform on their own that people who are specialists and Patrick was saying IML. You can do spectrum management, you can do anomaly detection, you can do other things around that optimization. And, so again, it's a new platform and the hope is that the R apps or the X apps, depending on which near real time or non real time, it's a new ecosystem. Again, more hopeful this time than the original SDN and SDN marketplace, which never really happened. But I think this time there are real world applications that we are seeing. And we've seen some of the work done like VMware with the Ric. With [15:03 inaudible] in terms of location, with Cohere in terms of doubling, improving spectrum efficiency. So there are now real examples. So I think I'm more hopeful this time than the original SDN one.
Patrick: I think Roy captured a lot of the key points, it's a segregation play, supporting an open ecosystem. I would comment on your meta. I think the meta is a long term bet. It's being driven by the technologists. I think it still needs to be sorted out, what's the clear path to get there and what's the investment look like to get there.
Abe Nejad: Well, so beyond AR/VR, again, put the training wheels on for someone in layman's terms beyond AR/VR, what is the metaverse or the metaverse network?
Patrick: Well, we were talking before. I mean, Matt and I were talking and you talk about digital twins. To me, digital twins is sort of okay, you can simulate and in some ways you can think of that sort of like the metaverse and it helps in terms of deployment. It helps in terms of testing and validation before you actually roll the service out.
Abe Nejad: But digital twins has been also been around for…
Patrick: It's been around and it's still being leveraged. It's early phase. The term's been around, but I think in some ways it's like 5g. We talked about 5g 10 years ago and now it's finally here. So the metaverse and how that fits into the ecosystem of the telecommunications, you're going to match up capabilities. We'll be talking about 5g. We'll talk about slicing. We'll be talking about eventually 6G and what 6G brings to the table. But in terms of how do you make money, in my opinion, you're a long ways away from monetizing the metaverse. You may be five years away, maybe longer.
Matt: The digital twin analogies is quite a good one, except there's kind of an extra layer with metaverse, which is interaction with said thing. Whereas digital twin is more of a representation, I guess. And then it's about interacting with that, which then theoretically needs lower latency, probably high bandwidth. So you can see why the telcos are rubbing their hands with glee, because this is pure telco stuff, great. Everybody needs, every home will need a 5g connection because you need real time, ultra low latency, you need however many gigabits second of connectivity.
Roy: 4k, 8K, 16 K
Matt: Exactly. So you got to have that. So everybody's going to need multiple lines and so on. So you can see the appeal there. I mean, for me, it's an opportunity to interact with customers. I think about this stuff from the enterprise standpoint, rather than from the consumer standpoint, but from the enterprise standpoint, if you are a retailer, great, you can have a store that you can sell to people. There's opportunities there, but I think we're still ways away.
Roy: We're ways away. I mean, second live did it prematurely. And now you may have a second live, a second live. But I think for me it's more about digitization. Digitization of people, digitization of items of devices, digitization of processes and moving all that into well, the term is metaverse I wish they were a better term for it, but it is what it is, but it's the digitization of those things. And then the ability to work with it in the digital realm for efficiency, for engagement, for simulation, regardless of whether it's people, whether it's devices or objects or whether it's processes. I think that to me is interesting enough value and to move the data in and to interact with the data, it's a huge amount of data. And so the networks are critical, the computing is critical to make that happen.
And so you have as Matt was saying, the service providers are [19:20 inaudible] excellent I can transfer more bandwidth and you'll pay for it because there potentially is hopefully business value or at least entertainment value in doing some of that. And the compute guys are really happy. Videos of the world are really glad because you're going to have to do a lot more computing, specialized compute at that. So for me, it's a digitization of our physical wall in there. And again, I wish it was a better term, but it is what it is, it's the metaverse we'll live with it.
Patrick: Yeah. I mean, you've got Microsoft doing some acquisitions in the gaming space. Do they become the platform or is it Meta, Facebook becomes the platform? You're going to have many different platforms. So this value chain and where the telcos fit in there is going to be something that needs to be sorted out.
Roy: But more bandwidth. Better performance.
Abe Nejad: Let's talk about, you mentioned acquisitions. That's probably a good place to finish. Certainly some news in the last several months about acquisitions, one being the Altiostar, Rakuten symphony acquisition of Altiostar. Now they're called Rakuten Symphony. And of course this morning the acquisition of Robin.io by again, Rakuten Symphony, that was sort of big news as well. Roy, if you don't mind, I'll start with you and then go around the room here. Where do you see the natural evolution of the conversions of our industry? Do you see more of the same of what happened this morning with Robin?
Roy: So I do see that having the underlying cloud platform, some of the orchestration elements, which Rakuten Symphony already had some of those elements in there. I think it does make sense in that, telco has specialized workloads. And to the extent that they can control a vertical stack of those elements, even though it's Kubernetes. There's a little real time capability that is IO centric capability that's required. And either you get someone like an AWS or GCP or Azure to build it for you, which they are right. Or you convince VMware and [21:19 inaudible] to do it for you, which they are, or you have it in house, like in the case of Rakuten Symphony, they've decided that makes sense. And so I think that aquisition is one strategy, have a full vertical stack where you control all the elements to get to market and you can usually get there faster and it's more specific.
And I think that works for them. And I think the question will be the rate of innovation in that field versus a more horizontal platform that's sort of being pushed out there by some of the players. We'll see how that plays out in the long term, because I think those are sort of doling strategies, both of which are very feasible. And I would say again, they've made very smart moves. Rakuten Symphony has made very smart moves over the last couple years. We look at them and we learn from them. So I think that's a good move so far.
Ade Nejad: So Matt, you, and then Patrick will finish.
Matt: I have nothing to add to that. I think that that summed up mainly.
Patrick: So I think they're also, they may have come out today in SIM world branding. And they want to provide a set of modular solutions. They can sell individually with respect to the moves that Symphony's done. I mean, I think they've been very strategic. So they've had an investment in every one of these companies that we're talking about, Altiostar, Robin.io. They're basically saying, okay we've kicked the tires, we're convinced, so we're going to bring them into the portfolio. I think any acquisitions that we see probably over the next few years, you're really going to need to justify it. And it's mostly because money is not free anymore. You're in an environment where there's less money going into the economy, quantitative tightening. And so you gotta make the right sort of bets, if the acquisition meets the criteria in the portfolio. But I do, I think in our sector we will see more acquisitions. They'll be carefully, thought out. Less maybe reckless than what we've seen in the past where you had free money.
Matt: Well, maybe potentially more strategic acquisitions rather than B VC, which is money going in.
The whole stack things.
Roy: That was a mess. That is a mess.
Abe Nejad: Great. Okay. Well, I'm so glad that you guys had the time to come in and do this. I think it's important to sort of wrap the day or even wrap the week. So if we can't have you back later in the week when everyone schedule sort of tight maybe we can do it either virtually or at another event down the road. So Matt, again, if we haven't done this before, glad to do it, we have done this before many years ago, which is very possible and good to have you back. Roy, we had a number of virtual opportunities over the last couple of years, but it's so great to see you in person, was looking forward to it and here it is. And Patrick we have not done this before, so it's good to have you here and certainly have you back soon.
###
For any inquiries, please email anejad@thenetworkmediagroup.com