Converged or divergent. Why your network might not be what you think
Converged Network Services (CNS) is a well established acronym when designing and delivering networks and are usually discussed with confidence by marketing, designers, developers and integrators alike. So is everyone comfortable with the approach and concept? Maybe not. Perhaps your network is actually just shared or in some cases a “divergent” network? In this opinion piece I wanted to share some of my own perspectives.
The Oxford English Dictionary definitions
The use of the word “converged” in context of an all encompassing network that unites all components is actually something very different. Often networks are simply shared resources of at worst “Divergent”.
For context, some integrators define convergence as : “The network running through the building and connecting the building systems and software, to which the occupier can integrate (including the physical environment and cabling).” The key part here is “can” and implies a level above simply sharing a network. So how many people actually achieve this integration?
There are very few statistics available for the amount of converged, shared or diverged networks. It’s very subjective in approach and important context is required before firm conclusions can be made. Try looking for statistics on the amount of converged networks deployed in the UK. The majority, if there are any, will point to public broadband convergence and not in building networks.
When you move away from looking for research into convergence, we move into the world of marketing and product sales. Here the field is clearer. Suddenly all networks are converged and capable of being so either through a complete product stack and licences or via expertise of an integrator but without any hard and fast definition beyond the art of the possible. Be it DNA Spaces from Cisco or the impressive Aruba CX platform which are extremely capable of doing everything that convergence could bring, very little of the case studies you will find online discuss the complexities of true convergence. By this I mean the difficult parts of integration through segmenting data, network authentication, sharing 3rd party networks and the choice between IT and OT networks which completely blow convergence out of the water.
Only 37% of deployed networks achieve full convergence
Over the past six years, we have designed approximately 50 network projects and have looked back at the component parts of the designs we have produced. What we have seen is that whilst over 80% of networks share all connected network elements, only 37% of those designs achieve what can be truly be described as achieving “Convergence”. Given the very small to non existent research into this available, I would encourage others to share data that they have and given the amount of publicity around “Converged Networks” this should yield some interesting comparisons.
Mundial Consulting Data taken from 50 designed projects between 2017-23
Note that I’m being careful with the difference between sharing network resources and those which are truly converged. A distinction here and a personal view, Smart Networks do not inherently converge a network. They require one.
The other side of the research is that of those networks which are designed purely for basebuild / landlord only 17% are converged with 83% not having any convergence at all and in some cases a network does not form part of an instructed package.
Question then firstly, why are fully deployed building networks shared in the vast majority, but when discussing basebuild networks, only 17% of these are using converged networks as part of their instructed packages. Secondly, of those who did share their network infrastructure, why were only 37% of networks fully integrated and converged? The other way to look at this is that 83% of basebuild networks or 63% of all networks who could have or think they have converged networks are actually divergent and in most cases by choice.
The answer should lie in whether the network is actually being designed and procured to a set of customer criteria or just because a project needs a network and instructed via disparate procurement activities. We personally find a lot of resistance from main contractors not wanting to take the risk of delaying a PC date to achieve the best outcome for a project or project design teams who don’t understand the need for convergence and only see cost. Either way what is described as a converged network is actually only present in just over a third of networks which paints a very different picture to the perception of Converged Networks being the norm.
So why the talk of convergence or the belief that networks are truly converged when they are more than likely simply shared infrastructure?
Who gains and why? At lot of this can be said to be driven by marketing and product sales. The lack of peer reviewed research (mine included and I would invite those who design to produce their own statistics) into this subject suggests that talk of convergence is subjective. (Our figures include a significant number of FTSE100 companies and are between 50,000 and 1,000,000 sqft for context).
So Is talk of convergence therefore simply marketing hype or a way of bundling licences on behalf of manufacturers and sharing a network without the difficult conversations being had with the end client? In this opinion piece, No and Yes. Convergence is absolutely achievable and is evidenced in over a third of our research however the gap between this figure and the 80% who may think they have a converged network points to misunderstanding what is actually required or installed. I won’t call out network manufacturers here but clearly the gains available in the marketplace are significant to those adapting their products for convergence and it is for us consultants to fight the corner and place the importance on the designs produced.
So how do organisations, design teams and customers stop themselves from falling into this misconception? It requires an independent step back from the product marketing to balance internal IT and customer teams against sometimes mis-aligned project timescales and to bring a laser focus on the What, Why and How. By increasing the level of integration and understanding, additional benefits to customer experience and building operational effectiveness, can match or exceed the investment benefits of shared infrastructure. (This will be the subject of a further article).
Think about what is important when designing your network
What are the customer requirements and benefits
Why are they important
How can these be facilitated by a network.
Are competing priorities or previously instructed works, diluting the original concept for a building?
When speaking to customers I always keep in my mind or likely say it aloud “IT should not be procured for IT’s sake” as it becomes self serving and self defeating in one move. Networks should rarely be seen or heard but their impact profoundly felt as an enabler for wider benefits and by focussing on these points, the discussion around convergence will take its natural place in a project, which is to facilitate a customer or business objective. Converged or not, it will then be delivered by design and not purporting to be something it isn’t.