Well, hello, everyone, I'm really delighted to welcome you all to TNC's latest podcast on the topic of "Is Co-management the Future Model for your Network?" I'm John Waterhouse, CEO of The Network Collective, and I'll be your host for the next 20 minutes. As I'm sure everyone joining knows, TNC is the UK's largest independent telecoms strategy and sourcing consultancy, supporting over 280 major UK and multinational companies to get the best commercial, technical, operational, and contractual results from their network and telecoms solutions. And joining us today to share his expertise is our CTO, Craig Northveth - Craig, would you like to say hello to our viewers and listeners?
Yeah. Hi, everybody. Welcome back to the Podcast. Hi, John: nice to have you back for this one as well.
Thank you very much. Glad to be back - it's a slightly hotter day than when we last recorded. So this is a really interesting topic today. Not least Craig, because you and I wrote a White Paper about this a few months ago. Yeah, so great to get an opportunity to revisit it and to do a podcast about it - and of course, it is a really interesting topic in itself. So as we've been reporting for some time now, through lots of podcasts and lots of articles, there is this technological revolution sweeping through the enterprise network market. However, harnessing sexy new technology is only part of the story. So for organisations to achieve their objectives of delivering magnificent agile, flexible network services, it's not enough just to go after the sexy technology, you're also going to have to build an agile, flexible operating model to support and deliver that infrastructure.
So we're going to talk today Craig, are we not, about Managed Services; where the world is today; but crucially, what are these new models that are coming along and coming into the market to help organisations to really gain the benefit out of the new technologies that they're deploying? And I have no doubt, that we're going to start talking about Programmable Networks, and Intelligent Networks, and all these other sorts of things - so pretty interesting topic for today. So let's get straight into it - you know I love flinging some questions your way - so could we start off today with just an overview? What is the "State of the Art" today? What are most organisations (and most people listening), likely to be doing in terms of managing their networks?
Yeah, I think it's a really good starting point, John, I think a lot of people listening, and obviously, a lot of people that we interact with, (through these types of sessions or breakfasts, or just general meetings), tend to be coming from a managed service environment - a fully managed service environment. This traditionally would be a single vendor, or a small number of vendors - managed service providers - effectively delivering the full end-to-end support and infrastructure and software across their entire network. So when we talk about the network as well, that could be within the wide area network - could be the local area network, could even be some of the contact centre services - traditionally, a lot of people have opted to go down that kind of fully managed service route. So they've got, you know, one single contract, one single support service, one single SLA effectively, for the whole end-to-end piece. So I think...
...Sorry, we're talking about that full stack of services, right? So, the network access, backbone, CPE, management, the whole shooting match in one in one contract?
Yeah. And I think I think the reason that's the case is that traditionally, these things have always been interlinked. So you've always thought of the Wide Area Network being the access, and the CPE - the Cisco, or Juniper router that sits on the end of it, and then a demarcation point there into your own internal network. I think as we move on and talk about it in a second, is we're starting to see these things slightly differently. Now we're seeing some separation between the different components, which is starting to now drive a different way of thinking around how to manage the network, and how to get the best flexibility and agility out of the kind of technology that as we talked about, is evolving and rapidly coming into the market.
So I mean, that's a perfect sort of entrée into the second question really, which is: OK that's where most organisations have started their transformation journey, what is it about that - let's call it the traditional model, which may no longer be fit for purpose, that's making organisations look at what they might want to do in the future, that's different?
Yeah, so there's probably two or three areas that we can touch upon here. I think - of the first two, if we start with start with technology, I think the evolution of technology is allowing - let's say - the simplification of management. So, if it's within the Wide Area Network, we're seeing things like SD-WAN capabilities being able to be centrally orchestrated through a web based platform: we're no longer having to have the complexities of command line configuration to set up a site, or make policy changes to the site, it's done quite simply through sort of 'drag and drop' type capabilities from a central point. So the entry point effectively to manage the network has become much, much simpler, it doesn't necessarily require the intelligence and capability of a managed service provider to do. So I think that's the first point.
I think the second point is, where we've talked about this separation of technology now as well. A lot of people, particularly within the WAN space, look at WAN access and WAN Edge infrastructure as two separate entities now, because WAN Edge infrastructure has become far more intelligent. We have started to see much more consolidation: convergence of things like security services, or even virtual CPE and many other network services from there, which don't always naturally sit with a managed service provider, like a telco. So we are starting to see that separation - organisations saying, "Well, we'll keep our telco/our managed service provider as providing access, but we want to take back over control of that overlay service, and certainly, the intelligence and complexity and agility that sits within there.
I think the final point to call out here, and it's always an important point for us, is cost as well. So, you know, wrapping everything up into a single managed service provider tends to be quite a costly approach. You know, we see various different price points in terms of the overall percentage of cost, which is made up of managed service support from an MSP, and it tends to be quite high. And you know, the lack of flexibility that you get sometimes with that model - people are starting to consider, "Could we do this better ourselves?","Can we do it cheaper ourselves?", or "Is there an alternative approach to do this..."
And that point about, you know, "Can we do it better ourselves?", I know, one of the points we picked out in the White Paper was - at least the perception from end user organisations was - that sort of single provider model, the traditional model is a bit slow and a bit clunky. Is that a fair comment?
It is when you compare it to the potential capabilities that the technology delivers now. So you know, a good example of this, and we've used it with customers before is if you're going to buy into the kind of technology evolution buy into the capability of having the agility and flexibility that offers, if you then put a managed service provider wrap around that, they tend to still work in a fairly traditional model. And so you effectively lose a lot of the benefit - the technological benefit - by overlayering a managed service provider in there. Unless, you know, I'm not saying everybody's the same; there are some out there that are certainly better and we have started seeing different people coming to the market, like service integrators, which are standing up services predominantly to support this type of technology. But yeah, a high percentage of the vendors out there will still act in a slow and clunky manner to get changes made, and yeah, again it's probably fairly costly as well.
So given we've talked a lot about the drive for agility and flexibility in networks obviously, then leading on to leveraging new technologies, whether it's SD WAN, or we're talking more and more these days about SASE, all these sorts of different models, they're about achieving an agile network. I guess, from what you're saying, you can start to see why the traditional managed service model perhaps isn't that sort of optimised model for the future? So I know, through some research you've been doing, you've looked at three potential future operating models to try and address some of these issues. Could you take us through those and how those alternatives might look in the future?
Yeah, sure. So I think I think the first one, and this one's probably been around, it's been the direct competitor to the fully managed model for some time, it's the In-house Management model. So you know, you can think of this as "Out-sourced versus In-sourced". We've seen some organisations effectively take this route, particularly with the kind of 'off the shelf' technologies like Cisco Meraki, for instance, and then going out and buying local internet services, and then having an element of success with that, you know, from a commercial perspective and from a speed of deployment, or general management point of view. But we do also see some complexities with this model, in that the more service providers, the more kind of local service providers that you add to the mix, the more complex it gets in terms of service management, billing, management, contract management, vendor management, because you know, it can get quite complex quite quickly. So although there is a place for elements of 'in-house management', particularly around where you can add some intelligence and agility to the overall network design and network management, going full 'in house', is something that we've seen some organisations struggle with, just because of the complexity. I think we've got examples where some customers have decided to go down that route, but then decided to pull some of it back into a managed model because, of it is in a "too hard to do" box. So you got to find the right balance with that one, I think.
And presumably, in that you've got a lot of kind of low level challenges in there ref having sufficient resources, having the right skills, keeping people trained, you know, there's lots of quite kind of detailed challenges in there, which, traditionally, organisations haven't necessarily waited to grab onto.
Yeah, I think so it's a really good point, the kind of acquisition and retention of skills in this area, and particularly, as that starts developing out into new technology areas as well, can be difficult. You've got to think, to try and replicate the resiliency and sustainability that the MSPs have got, to replicate that internally, particularly if you're like a 24/7 business, this is fairly challenging to do - and it can be fairly costly as well. So the model has got to be right, the environment has got to be right to be able to support that 'in house' type of approach.
And the second option which is a relatively new one, (and we are seeing some of the MSPs actually offer this, but we tend to see this more in the SI space), is this concept of 'Build and Transfer'. So this is where, a service provider would come in to effectively stand up the whole infrastructure, stand up the connectivity, stand up the contact centre, or whatever technology it might be, and then transfer the orchestration of that, and the support of that upon completion. This is quite attractive for some organisations where they've effectively got a pool of BAU resources, but not necessarily a large pool of Project, or Development resources to effectively design, and develop, and deliver, this type of capability. So we are seeing that becoming - I wouldn't say massively popular - but we are starting to see quite a lot of interest in that capability, to do that and then hand over the support either to their internal teams, or more often than not, to another outsourcer that might be in the mix, some sort of, you know, business process outsourcer, or SIAM-type outsourcer, giving the keys to them once it's effectively been built, and delivered.
That's interesting, so you get the specialist resource for the kind of the high intensity activities around design, architecture, deployment, etc. But once you've got that steady state environment, you hand that over. That's interesting.
Yeah, so I think that will be one that we will see a bit more of in the future. Because again, tied to the things around the limitations of the 'in house' model skill limitations, you've been able to acquire and retain highly skilled people to do some of this design and deployment activity. It's quite easy just to come and pay someone as part of a professional services bundle, get that delivered, and then you transfer it back to the in house resources, which over that period of time you can train up, or you can give it to an existing incumbent service provider within the organisation.
So I think the final model, and this is the one that we're probably seeing has been the most popular, and this is across the board again, from an MSP and an SI perspective, is this co-managed model - the concept of co-management. This is where we're starting to see certain capabilities and functionalities being handed back into the customer base, so that could be simple policy management, routing changes, anything that can be delivered through the orchestration platforms or web platforms for things like change, is being is being effectively handed back into the customer to try and enhance that flexibility and agility of the management wrap. So this is where we see the best of both worlds really, so you've still got a good solid, single contract or multiple contracts, supporting the overall infrastructure in the network but you're also allowed to, you know, you've got the keys to go and think yourself as well, and make changes that are critical to your business but do them rapidly, rather than having to wait for the process of the MSP...
...and I was going to say presumably, there's an element here of trying to sort of have the best of both worlds - that the customer has the agility when they need it, as you say, quick urgent changes required to respond to, a business impacting situation or whatever, but BAU can continue to be outsourced and managed within that sort of contractual environment with a service provider without the need to resource it yourselves and so on?
Yeah, absolutely that. I think there's downsides to this model, and what we tend to see is that commercially, you would think because the responsibilities have been shared, it would be cheaper than a fully managed model. I think in reality, it probably is still a fully managed model, but with some flexibility to allow the customer to make changes. So it's still priced, (and this, what we've seen so far), is still tends to be priced as a fully managed model but ywith the ability to make some changes without having to go through the MSP process, or SI process. So commercially, there's not any real benefits that we've seen - so far. It's more around the kind of productivity and operational benefits, in terms of being able to manage the network.
But it's trying to get that balance of - and I can see the customer logic, that very often, the thing that sticks out to customers and stands out to them when they're thinking about their management models is, "I remember an issue when such and such a thing happened, we needed to be on it and we knew what needed to be done, and our service provider took 24 hours to make the change - we could have solved that there and then", but whilst that is a great case for 'in house' management, it's not actually a great argument for long term 'in house' management, because it's a one off thing, most of the other things that happened in the background are probably fine. So it's a way I guess, of combining those two.
Absolutely. I mean, I think that's a great example. I've personally had that experience in my life as well - you can see what the problem is, but you can't you can't get it fixed, because it's locked down, you've only got "Read only" access to certain things, and it's very frustrating. But then, on the opposite side of that is - "Would I want to be called it for four o'clock in the morning to fix every single problem? Probably, I still want somebody to manage that service for me and make sure that 24/7, we've got that capability regardless of whether I'm available or not, or the team's available or not. It's a good example. I think we'll see more and more of this, and what you can do as a customer will probably increase over time, and I think we'll also start seeing much more automation being built in some of these processes as well. So that co-managed model one, we've got the orchestration that the customer can do, of being able to maybe integrate that into some customer automation platform. So it could be tied into a service now, for instance, and start making changes that are automated and do not require that kind of human interaction, such as issues being reported. So I think we're gonna start seeing more and more of that going forward.
Yeah. But presumably, I am just starting to think about some of the main pros and cons for these different models, and we've touched on on a number of them, as we've gone through but, just thinking about, perhaps if we start with the co-managed and work back through the options, I can see those benefits of the flexibility, the agility, whilst removing the complexity of having to design and build and manage your own network, presumably, then - you've talked about the downside of the cost - presumably though, one of the downsides of co-managed is you're still limited to the technology capability or the technology solution set offered by your service provider?
Yeah, so you are. I mean ultimately, you still need a service provider in the mix with this model, and you are limited by their both their technology, availability and their capability as well. So you know, you can supplement that with your own internal capability. And potentially you can negotiate the level at how that co-management works, the key service boundaries, but you are limited by that, you've not got an open catalogue, effectively being able to select what technology you want...
Whereas presumably, (which to some extent must also be the case with the 'build and transfer'), obviously the benefit potentially, with the in-house managed is, you can go and buy whatever you like because it's your problem to manage it. So as long as you're willing to manage it, you could go for any service provider, any vendor, any combination of services, etc. because it's going to be your issue...?
You could do on the 'build and transfer' to a certain extent, but then particularly if you select a certain type of technology, you'd be limited to the specialisms within that technology, within the market, and who could do like that kind of service for you. So I think you know, the more you get away from the norm, the more likely it is you're going to be looking at a kind of bespoke or boutique type of professional service outfit to be able to do this. Which isn't, you know, not always a bad thing. I think, particularly when we look at some of the agility and flexibility of working with someone outside of the big, MSPs out there, people are starting to consider this now as well, and, you know, seeing the benefits of that. So, it does open the market up quite significantly.
Yeah. So, I mean, just looping back to the topic, what we set out to talk about, you know: "Is Co-managed the future model for for for network management?" I can see where that thought process comes from. Insofar as, the other models, the managed services, well, we know what the problem with that is, - limited technology choice, it's potentially slow, clunky, it's inflexible, and of course, it's expensive; In-house managed, well all right it's flexible, you've got a wide technology choice, but Cor Blimey! A lot of organisations probably don't really want to skill up, and staff up, and resource up, and technology teams? Yeah, that's hard to scale, Hard to build, potentially lacks processes etc. 'Build and transfer' solves some of those, but you've still got to deal with that. So I can see, I can see why Co-management ticks a lot of boxes. We used the term earlier on, 'best of both worlds.' You know, it does feel like it has that element of the security of a Managed Service, but it also has the flexibility of that sort of In-house model? Is that where you see the market going over the next 1, 2, 3, 4, 5 years? Do you think that's going to be the winner?
I think it'll be driven driven by two factors, really. So one, the MSP approach to the proposition to the market I think will start swinging that way, because this is probably seen as a "Value add". We can give you access to make these changes, and you can have a little bit more control over what's happening in your environment, we're just going to provide the the transport mechanism and make sure all the lights stay on and everything's fed and watered, but we'll let you do some more stuff, and that's, as a value proposition, it's probably quite attractive. The challenges, and you called it out right at the start, is that would require a different operating model for the customer. So it will require some investment in people, and skills, and capability,within the customer base to be able to actually consume that as a Service. So I think the market will start shifting that way but it will also offer, because it is best basically still dressed up as a fully managed service, they'll still do everything, or they'll let you do some stuff - or you decide not to do it, you do everything yourself. But it will come down to the appetite around wanting to change the operating model and take more of that service in house. I think for some organisations, we are starting to see some of this because it's happening in other areas like Cloud computing, for instance, and in Security. And we're starting to see people build kind of DevOps type culture, where they've got more ability to make changes themselves to their own environment. I think as then things continue to become more and more integrated and you're kind of designing things with security and network in mind, from an application point of view, it'll naturally start moving that way. But the technology's probably not quite there yet to enable that to happen. But yeah, I think the general direction of travel will be a slow, swing more back into giving the customer some more control over how things work, and yeah, obviously, it's the next phase that we've talked about in the past - programmable networks, thinking of it as an API almost, then that's definitely when it'll start being factored into more of an in house or development type capability, rather than a traditional MSP type capability.
Fascinating as always, Craig, and as always it feels like a shame when we reached the end of our time, where we've got to wrap up because I know we could easily keep talking about this, and there's much more to talk about, but we have to draw it to a close - our 20 minutes is up. And yep, we'll have to save any further chat for our next podcast. But Craig thank you as always for providing us with your extremely interesting insights.
Thank you, John.
Thank you everyone for listening and watching. As always, please do let us know any questions you have about this, or any other network or telecoms topic. You can get in touch with us through our website @ networkcollective.co.uk, or any of our usual social channels. We look forward to talking with you again soon.