We have been meaning to touch on this subject for a while, it is important, but apologies upfront, there are going to be a lot of acronyms. What better way to start the week than with a good acronym. These are interesting and could end up being very important.
Let’s start with WAN – a Wide Area Network – the network for an enterprise with multiple locations – which is pretty much every enterprise today – this is the entity we once called the corporate network. How do companies connect all their offices? This was always a tricky problem, and it has gotten much more so with the advent of widespread work from home. If all of a company’s employees are in one place, these are pretty simple. Wire up all the PCs and cash registers, through a switch in the closet, maybe some Wi-Fi, and that’s a corporate network. But once they add a second location, it gets trickier. Now the company has to rely on someone else’s network, often some combination of the public Internet and a telco or other service provider’s gear. This is a WAN.
For a long time, every company took a different approach to this. It was a headache, but one largely confined to the networking professionals on the IT team, They were the only ones who really had to worry about it, until it went down, or until the CFO got a wild bill from the intermediate service provider. Out of this problem emerged a new category of network equipment providers. And about a decade ago, this area took on its own name – SD-WAN – a Software Defined Wide Area Network. There were still lots of ways to build a WAN, but a class of companies emerged to manage this mess. At first, these companies sold networking boxes – the enterprise would plug one box in at the head office and one in the branch office, and the box would manage traffic between the two. Eventually, this function morphed, and companies stopped selling hardware and started selling a service. The head office would send traffic to the SD-WAN provider who would then direct that traffic to the branch office, removing the need for the company to install any more hardware. As with all things in networking, there is never one solution, and while we make this sound like a seamless process, it was instead very fragmented with many different companies selling different versions of SD-WAN.
For buyers, there are several benefits to using SD-WAN. The first reason they bought these solutions was to reduce their bandwidth costs. The SD-WAN can send data along multiple paths and can be programmed to direct the traffic to the cheapest path of the moment. While this is still a factor, SD-WAN has other benefits which have gradually become apparent. For one they can improve the network. To give the most basic example, instead of directing traffic along the fastest path, they can direct to the most reliable. And here the terminology gets a little muddy as there are many ways to do this. Once upon time, this sort of application acceleration was seen as a separate category, and while many companies sell that sort of solution, there are still a dozen or so SD-WAN companies that treat this as just one of many features in their tool offering.
Along the way, the industry consolidated. Cisco acquired Viptela. Riverbed was acquired by private equity. The list goes on. Normally, that would be the end of the story. A few start-ups pioneer something new. The big companies either acquire them or build their own solution, and then the whole category becomes just a set of features in some BigCo’s product suite. But this did not happen to SD-WAN. Instead, the category moved up the value chain. It was around this time that the providers stopped selling on bandwidth cost savings, an area that was becoming a table stakes commodity, and started selling on all the other features they could provide. The independent companies survived and flourished. Today, the category remains very healthy, this is still a multi-billion dollar market.
And it could become much more interesting in the future. There is obviously a lot of change taking place in the way that enterprises large and small run their IT systems. Even setting aside work from home, there is still the transition to the cloud. With companies no longer hosting all (most, any) of their own compute, they still need to manage and control who has access to their data as it runs through the public cloud. This makes for a very complex management environment, there is a lot of code and a lot of data moving all over the place.
The remaining SD-WAN companies see this as a major opportunity. Since they are positioned on top of their customers traffic, why not help manage that problem? This really boils down to two terms we often see in the SD-WAN companies’ marketing materials: Orchestration – keeping track of and coordinating all the traffic; and Security controlling access to the traffic. This is a fairly dramatic re-envisioning of the network, so dramatic it merits its own acronym – SASE. Let’s not worry about what it stands for, the term itself has gotten to be somewhat controversial, but it conveys the idea of an SD-WAN provider doing more than just managing bandwidth. These are critical functions for enterprise IT teams, and they are reasonably reluctant to give a single vendor too much control here. That being said, a lot of companies may see value in having someone help them manage their networks in this way.
If you squint and use a lot of imagination, it is possible to see SD-WAN providers becoming very important in the network. If they have control of network security, a massive market, and in managing the network, a massive pain point, this positions them to dominate the way networks get built. Now before we get carried away, it is very important to remember that there is never a single solution to networking, no perfect way, no standard. There are lots of choices all with their own trade-offs. There are also some massive companies involved who have a very large vested interest in seeing things do not go too far down this path, unless they are the ones steering.
Regardless of the pixie dust future, the present for SD-WAN companies is already pretty good. Even prior to Covid, there were some strong forces boosting their market, and work from home has just exacerbated many of those forces. This is an interesting, and possibly very important market.