Ray Le Maistre, TelecomTV (00:12):
So we're at Future Net World 2024 in London. I'm here with Francis Haysom, principal analyst at Appledore Research. Francis, you have been really involved in the agenda here, up on stage obviously in a lot of the sessions. Overall, what's been your main takeaway? What has surprised you or enlightened you at this event?

Francis Haysom, Appledore Research (00:34):
I think the biggest, biggest thing is actually this show grows from strength to strength. I've been at this event since its first one, I think back in the Inmarsat building, before Covid.

Ray Le Maistre, TelecomTV (00:46):
Old Street.

Francis Haysom, Appledore Research (00:46):
And I think it's always had this wonderful dynamic of a realistic agenda between people are taught, not afraid to say the difficult subjects and the things that work and are not working and to have a dialogue in the industry. And this one, just like last year, and the previous ones has kind of built on it. It's much bigger. I think it's approximately one quarter, one and a half times the bigger as last year, which is great. And the dialogue's been great. Some old friends here as well, and also some new entrants here. As you know, I was at Q con a month ago and it's great to see some more of the kind of software development, software development places, not just Red Hat, but now GitLab as well coming into the fold here. So that's a good sign. So conversation's been conversation has been great. Lots of dialogue, lots of CSPs here. That's actually the key thing. And lots of CSPs not just at the C level, but at the kind of doing level and operational level as well. So that all good.

Ray Le Maistre, TelecomTV (02:07):
Now, a lot of the conversations here have been around automation and of course the use of AI in that. It's very noticeable here that a lot of the companies from the vendor side of the fence are the kind of companies that we would normally expect to see at TM Forum events. They're companies that have come from the OSS and BSS world. And for me, I think that's really good as well. But have you managed to get a sense from this event about where that particular part of the sector is going? Obviously automation doesn't happen without the software to enable it. Is that software coming from that traditional telecom software sector or is it coming from elsewhere? Because I've heard more and more from more and more here as well that they're having to do quite or more stuff in-house these days. They can't get it from the vendor community.

Francis Haysom, Appledore Research (03:08):
I think there's a little bit of the answer as with all good questions. It's a little bit of both on that area. Definitely the vendors that are in the room here are really quite pushing the boundaries in terms of what can be done. As I've already mentioned, you've got the industries, other industries coming in from outside. I think the tension is actually quite a lot of it is do you see yourself as A CSP as being unique and you have unique requirements that need your unique solution. And I think there's always been that sort of tension between you to take something out of the box and use it or do you customize it to your particular, that's been decades old problem in OSS and BSS there. I think that the challenge is to recognize how do we start to use common features? How do we actually do both the CSPs recognize where they are really different and that they're really investing in where they are different and how do we see that they can use common capabilities either from their traditional vendors or from new entrance along the way.

(04:34):
So I think it's neither black one way or white completely the other way. The second thing I think is I think everybody's wanting to do quite a lot at the same time. You've got ai, the generative AI is the next best thing coming in here. And again, there's a lot of demands. I think as we haven't quite worked out, where does gen AI sit within this one? Many in the industry will be looking for standards to define what it is that they do. And again, that naturally lends itself to a kind of bespoke or customized way of doing things. And I undoubtedly believe the industry will start to see more repeatable methods that many of the vendors here might start to adopt. So I think we are in a kind of a difficult business. Lots of pressures on the industry, lots of things that need to change at the time, and that does drive towards a kind of customized view of the world. I think over time the market will mature. You'll start to realize actually I can buy a lot of that capability from the existing vendors or the new vendors. And also I think the important thing is actually realize the CSPs need to be innovating, but it's not necessarily innovating by replicating what vendors can bring them all today.

Ray Le Maistre, TelecomTV (06:09):
Right. And do you think in general, because we've been talking about shifting IT workloads, telco, IT workloads onto public cloud platforms for quite a long time. Have we reached a point now where that's just a given for a lot of these companies or is there still work to be done?

Francis Haysom, Appledore Research (06:28):
I think there's work still to be done there, but again, I don't think it, it's another technology issue, both a business model and a business thinking, operational thinking on it. I think the industry is still not, we are used to building networks for boxes and as we move to cloud native, and we've seen this with virtualization in a lot of cases, we recreated those boxes. But now as virtual virtualized boxes, cloud native means more than box. It means that a function that is here today could be over there. Tomorrow, could be a hundred over there in 10 minutes. It could be down to two tonight. And that's a very different way of thinking about the model. And I think that plays a lot into the way that telco seeds, public cloud.

(07:30):
But something in the industry say, look, public cloud is just far too expensive. And in one sense, they're right for a network that is very pre-planned known capacity, no number of subscribers that will be using a certain area known services, et cetera. The model of telco already knows how to capitalize that now capitalize as a data center or an edge data center. They're very good at doing that one. I think the public cloud, if you look at a lot of where public cloud has really taken off in other industries, it's where you don't know which services are going to be, you don't know the type of capacity you will need today, tomorrow you could have a service. It is the old TikTok. TikTok starts as a small thing and suddenly it's billions of billions of transactions. That thing cannot be done on a CapEx maximum size problem. That's where the elasticity of the public cloud comes into. So I think there's a lot of thinking that needs to go in where we're a steady state, we're probably going to end up as hybrid and private cloud because that makes absolute sense. But where we want, we are uncertain we need that and we're going to get a business benefit from paying. You are paying for that elasticity, but if you can work out what the business benefit that comes from having that elasticity, it's a very different business model, business success criteria.

Ray Le Maistre, TelecomTV (09:15):
And we've got quite a long road ahead of us in that. I guess the industry will follow in due course. Okay. Well Francis, great to talk to you. Good to see you.

Francis Haysom, Appledore Research (09:26):
As ever, Ray.

Ray Le Maistre, TelecomTV (09:26):
At this event. And it won't be long before we are in the same place again for a few other industry events.

Francis Haysom, Appledore Research (09:33):
Great stuff. Thank you, Ray.

Ray Le Maistre, TelecomTV (09:35):
Look forward to seeing you soon.