It’s that time of the year again when analysts and pundits flood you with their review of 2018 events and predictions for 2019. I’ll try to avoid the clichés and not devolve into a “here’s what happened in 2018” article, but instead provide insights for the next 12 months using information gathered from recent conversations with system integrators (SI) and service providers (SP). As you’ll see in the rest of this article, 2019 will be a year of opportunity for SPs, SIs, and Software-defined Wide Area Networking (SD-WAN) vendors, but a number of them might not make it the end of 2019. Here’s what I expect to happen in 2019 in SDWAN:

Further Consolidation

Just when you thought SDWAN news was going to slow down, Oracle purchased Talari. This comes on top of previous prominent acquisitions such as VMware purchasing VeloCloud, Cisco buying Viptela, and Riverbed purchasing Ocedo.

It doesn’t take much analysis to predict that more SDWAN vendors will either be acquired or go out of business because the market can’t sustain the 60 or more vendors that exist today. Beyond vendors purchasing other vendors, I anticipate that some service providers will enter the fray and employ a vertical integration strategy. We’ll see at least one SP acquisition of an SDWAN vendor in 2019 to cement ownership of its own SDWAN offerings. Whether that integration will be successful is a different story.

Erosion of MPLS

You might have seen statistics showing that MPLS customers haven’t gone away and that MPLS revenue hasn’t been hurt by the arrival of SDWAN. Remember though, that we’re at the beginning of the SDWAN trend and many SDWAN early adopters have simply left their MPLS in place while layering SDWAN on top. With vendors touting MPLS-like performance and improved service level agreements using multiple broadband links with 4G/LTE backups, there will either be price erosion on MPLS links or a reduction in new installations of MPLS, or both. This trend will show up in the data soon and is already being hinted at in my conversations with SPs.

Some SPs Will Effectively Sell and Deploy SDWAN

Service providers will continue to face the conundrum of either embracing, or pushing back SDWAN — or doing both at the same time. This is not unlike NFV vendors hedging their bets by offering lip-service virtual network functions, which may work but not very well, while still pushing their own proprietary hardware that may show better price and performance for now. Regardless, it’s clear that enterprises are massively interested in transforming their WAN, and business and competitive pressures make it impossible not to upgrade their WAN soon.

Some early service providers that have jumped into the SDWAN fray are facing challenges in selling the service. They are realizing a couple of things:

  • Their sales teams, used to selling basic connectivity services, might not be adept at selling SDWAN value-added services.
  • Their operational teams have similar issues running SDWAN services in their cloud infrastructure.
  • Their point of contact at the enterprise just expanded from the WAN engineer to include IT directors and even CIOs due to the strategic importance of SDWAN.

The lack of operational expertise may lead vendors to offer hosted or turnkey solutions that are white-labeled or co-labeled with the SPs. The recently announced Netcracker Business Cloud Infrastructure offering falls into this category as a managed, turnkey solution that the vendor operates on behalf of an SP. If this model succeeds, count on more vendors or system integrators offering the same.

SPs Need to Uplevel Their Partnership Capabilities

One of the tenets of SDWAN solutions is the redundancy afforded by multiple links. For this capability to be effective, SPs need to facilitate as much independence of the link paths as possible and ensure that cloud infrastructure termination point for these SDWAN links has sufficient redundancy. This means that a single provider should not provide all last-mile infrastructure. In addition to this, many SDWAN implementations will employ 4G/LTE links and will require SPs to partner with mobile providers, if they aren’t already a mobile operator.

At the same time, service providers will be pushed to offer fast connectivity to popular public clouds as part of their SDWAN offering. This means they either partner with the likes of Equinix, Digital Realty, and other Inter-exchange point providers, or they have to negotiate deals directly with the major public clouds.

All this means is that partnership capabilities for SPs need to be increased and their sales teams need to become adept at selling bundles of links, not just the typical VPN — L2 or L3 — offerings. It also means that SPs need to get smart about cloud infrastructure connection solutions into virtual private clouds in popular public clouds. Education and training for sales and operational teams will become a key element of success in SDWAN.

System Integrators Will Step Up

I expect system integrators to step up in 2019 and push managed SDWAN offerings into the marketplace. Excepting some of the largest enterprises that may choose to run their own global SDWAN offerings, many mid-to-small-enterprises will want managed services. SIs could also be in a unique position to offer cross-SP offerings, as well as create custom portals and workflows that certain enterprises need.

Because of their perceived SP-agnosticism, SIs could be preferred by some enterprises over a single SP. SIs have the flexibility to pick the best connectivity options from multiple SPs while building SP-agnostic infrastructure for cloud infrastructure-based management systems. They will also have more software-development capabilities to customize or extend the necessary SDWAN portals to meet certain enterprise needs.

The other aspect that could favor SIs is their ability to service enterprises beyond the narrow scope of the WAN. Given that SDWAN tend to span branch infrastructure, enterprise security, and cloud infrastructure development platforms, it might allow SIs to have a slight advantage over many SPs that don’t necessarily have that level of domain knowledge today.

Fundamentally, there’s nothing to stop SPs from expanding their services. Indeed, some SPs do offer value-added services in terms of managed branch, managed security, and even private cloud infrastructure hosting. However, there are many that aren’t used to selling more than just connectivity to enterprises. The conundrum for these is whether to stick with selling what they are used to and becoming more efficient, or to make the investment necessary to uplevel into a more comprehensive service provider with all the attendant risks. This decision will have to be made in early 2019 as the land grab for SDWAN worldwide kicks into high gear. Fortune favors the brave and bold, but also the wise and pragmatic. So which will it be?