Over the last 2-3 years, the marketplace for WAN services has begun to change, from a near-universal dependence on MPLS to a market offering a range of solutions which still includes MPLS, but where there are now many other options.

As in so many areas of IT, the movement to cloud-based applications is changing everything when it comes to WAN design.

It seems highly likely that this market flux will continue for some years while organisations change the way they deliver applications and vendors bring new technologies and solutions to the network and telecoms market to address these new application environments.

Below are two main approaches TNC sees to the WAN of the future, and the key characteristics and pros and cons of each.

Approach 1:
Hybrid and SD-WAN

One potential model for the WAN of the Future uses hybrid MPLS and internet, potentially also adding SD-WAN functionality for enhanced control and management.

For organisations taking this approach, the WAN of the future would look very different to the WAN of today.

Today’s WAN has three key characteristics:

  • Typically uses MPLS as the primary backbone technology
  • Tends to be delivered by a single or small number of suppliers
  • Usually delivered as a fully-managed service.

All three of those characteristics would change in a hybrid and SD-WAN of the future:

  • Uses a hybrid of Internet and MPLS in the backbone, with an SD overlay
  • May be delivered by multiple suppliers
  • Orchestration of these suppliers would be the customer’s responsibility, fundamentally moving the service boundary from where it sits today

Let’s look at these three characteristics in turn and understand what is changing and how this approach aims to deliver the results organisations require.

Approach 1:

Many organisations are finding that their MPLS networks are poorly suited to delivering cloud connectivity. This is partly because MPLS-based WANs can perform poorly for such connectivity, partly because the price points for MPLS make delivering additional bandwidth commercially prohibitive, and partly a desire to deploy services which get traffic to the internet more quickly, if not directly from each site.

For these reasons, many organisations are looking at alternative backbone solutions. Key amongst these alternatives is increased use of the internet, either as a pure-play or as part of a hybrid network using MPLS for some sites and/or applications.

Many organisations considering these types of solutions are also attracted by the potential for SD technologies to manage, control and optimise backbone performance.

Approach 1:
Multiple Suppliers

In the world of MPLS WANs, organisations tend to use a single or small number of suppliers to deliver their WANs. There are a number of drivers for this:

  • Interconnecting MPLS networks is expensive and can create a significant operational overhead. Therefore, organisations try their hardest to avoid doing so, which leads to a bias towards using a single supplier
  • Organisations tend to only break through this bias when there is a very strong commercial return. Given that local and regional suppliers tend to be cheaper in their specialist geography than the global suppliers, this commercial return tends to arise when there are enough sites in a particular country or region to offset the cost and hassle of interconnection.
However, a move to internet-based WANs changes this calculus considerably:

  • Whilst organisations could procure internet services from the same global vendors they currently use for their MPLS networks, very often organisations are looking at the price points offered by local or regional internet providers and realising they could make big savings by splitting up their networks.
  • Organisations are often frustrated by the slow response and reaction times of the global vendors. Given the strong drivers for more flexible and agile IT services, organisations often feel that dealing directly with local or regional vendors will enable quicker provisioning and more responsive service
  • The inherent ability for internet-based networks to interconnect means that creating and managing multi-supplier internet networks has lower commercial and operational overheads

In other words, instead of the bias being towards a single supplier in an MPLS world, the bias increasingly is towards multi-suppliers in an internet and hybrid world.

Approach 1:
Changing Service Boundaries

So, if we are looking at different backbone technologies and a move to a multi-supplier world, why does that mean we also need to look at changing service boundaries?

The answer can be summed up in one word: agility.

The drive to cloud-based applications and the wealth of new service architectures is driven by a desire for agility in all senses of the word – agility to deploy services more quickly, agility to turn those services up or down to match demand, agility to deliver change.

When organisations demanding agility look at the networks market they see the traditional vendors struggling to deliver agile solutions that leverage latest technologies at commercially attractive prices. At the same time, they see local and regional suppliers delivering cheaper services in a more flexible way, and sometimes exciting, innovative organisations coming into the market delivering new technologies. The challenge is that neither of these types of suppliers typically have the maturity, scale or reach to package these services into a comprehensive, end-to-end solution.

Many organisations are responding by taking on this role themselves – skilling up to integrate these services using internal capabilities, thereby delivering the agility their organisations are demanding.

In the traditional WAN world, people used to talk about two service models – “managed service” and “wires-only”. TNC sees this new service model as more of an internally orchestrated managed service. What this means is that each element of the service is managed by the supplier, but the end-user organisation is responsible for orchestrating these multi-suppliers into a cohesive whole.

However, the ability for organisations to take on these challenges is highly variable. For organisations that are already very centralised, the idea of taking on the management of multiple vendors, particularly in multiple geographies may be daunting. However, for organisations that are already more decentralised and have IT skills in different places around the world, managing multiple vendors in multiple geographies may seem perfectly reasonable. TNC’s research certainly does show that there is huge resistance for organisations to change from centralised to decentralised (or vice versa), so it does seem that the shape of an organisation at the start of its journey to the WAN of the future will be a significant determinant of the type of WAN and the service model it will find attractive.

It is also very important to consider the skill sets of the internal team. Most network teams are very comfortable managing MPLS networks with Cisco CPE, but few network teams have strong SD-WAN skills, given how new that solution is to the market. Therefore, migrating to a solution based on very new technology will be a risk, given the need to upskill internal teams, and of course retain upskilled staff who will look very attractive to other organisations contemplating the same solutions!

Approach 1:
Pros and Cons

TNC would summarise the pros and cons of the hybrid and SD-WAN models as follows:


  • Cheaper internet connectivity can allow organisations to deploy more bandwidth within existing budgets
  • Can improve application performance due to routing efficiency
  • Engaging with local and regional operators can increase agility

  • Internet services often have no SLAs so performance isn’t guaranteed
  • Utilising multiple providers creates operational complexity
  • Internal network teams may not have the skills to design and manage hybrid and particularly SD solutions
  • Using multiple internet providers can create security vulnerabilities

Approach 2:
Cloud-Connected MPLS

Another potential model for the WAN of the Future uses cloud-connected MPLS. Over recent years, many WAN service providers have been adding direct connections to major cloud providers from within their MPLS networks.

For organisations taking this approach, it allows them to avoid the inefficient routing to the Internet which can be an issue with MPLS, but whilst also gaining the significant advantage of having Quality of Service and robust SLAs underpinning cloud connectivity, which isn’t always the case with hybrid networks.

Cloud-connected MPLS has other potential advantages compared to hybrid networks:

  • Given most organisations already have MPLS, adding cloud connectivity may be easier than ripping out much of the existing network to replace it with internet connections
  • As it is an MPLS network, the number of vendors utilised to deliver the solution is likely to be lower, giving lower management complexity compared to hybrid networks
  • Also due to it being an MPLS network, it is inherently more secure

However, there are also a number of risks and issues to consider:

  • Many organisations have found that they can’t afford the bandwidth upgrades required to deliver cloud using MPLS, particularly taking into account the cost of cloud connectivity
  • Cloud connectivity is only advantageous if the clouds you want to connect to are connected by your service provider. That may be the case on day one but as new services come to the market, there is a risk that you will not have connectivity to the services you require

Beyond the
WAN of the Future

While the picture of the WAN of the Future set out above is already a stretch for many organisations, TNC can already see how even this model will change over the next few years. In particular, two key trends seem clear:

  • As quickly as MPLS gives way to hybrid and SD backbone networks, there is a question about whether organisations need their own backbone in the future at all. As local and regional internet connections pull traffic into the global clouds operated by behemoths like Microsoft, Amazon and Google, the backbones those companies already operate to manage their own traffic may become the de facto backbone for their customers, delivering optimised traffic to wherever it needs to go
  • Although the current trend is for service boundaries to move as end-user organisations take on more of an orchestration role, it seems likely that third party management organisations will develop into that space. Already some of the Systems Integrators are showing interest, as are some of the traditional carriers, so we may well see service boundaries moving again as new players enter the market, or existing players change their offerings


WAN services, as all networks and telecoms services, are going through unprecedented change. This has opened up tremendous opportunities for end-user organisations to radically overhaul how they deliver connectivity to their users, and to generate significant benefits for their organisations as they do so.

However, this change is also creating considerable risk and has made it much more difficult to develop and deliver coherent network and telecoms strategies.

Of course, that then begs the question – “which approach is right for my organisation?”

Here at The Network Collective, we help over 200 organisations make the right technology choices, pick the right supplier partners and deliver the best possible commercial, technical, operational and contractual results. If future WAN strategy is rising up your agenda, please get in contact – we’d love to talk with you.


Other than matters relating to The Network Collective, this research is based on current public information that we consider reliable. Opinions expressed may change without notice and may differ from views set out in other documents created by The Network Collective. The above information is provided for informational purposes only and without any obligation, whether contractual or otherwise. No warranty or representation is made as to the correctness, completeness and accuracy of the information given or the assessments made.

This research does not constitute a personal recommendation or take into account the particular investment objectives, financial situations, or needs of individual clients. Clients should consider whether any advice or recommendation in this research is suitable for their particular circumstances and, if appropriate, seek professional advice.

No part of this material may be (i) copied, photocopied or duplicated in any form by any means or (ii) redistributed without the prior written consent of The Network Collective Limited © 2018

Let's talk
About how TNC can help you optimise your network and telecoms solutions
Lets talk

Call us, email us or fill in the form to talk to one of our experts. TNC is the leading network and telecoms consultancy, we are here to answer your questions and share our unique industry insights.

By filling out this form and clicking ‘submit’ you agree to our privacy policy and agree to us processing your personal data. You can opt out at any time by sending an email to or clicking unsubscribe in any of our communications. 

[javascript protected email address]
Let's Talk