Every day I talk to people who want change. They are disenchanted. No, I’m not talking about the current political climate. I’m talking about the companies I speak with every day about their WANs. The stories I hear have as much drama as we see watching the day-to-day ups and downs of the White House.
The companies I talk to are primarily in the middle market, in vertical industries ranging from retail to healthcare to finance to a range of professional services. The mid-market represents a huge slice of the U.S. economy, but it tends to be the overlooked middle child in so many ways: Not as big as deep-pocketed large enterprises and not as numerous as small businesses. The mid-market tends to be under-served in a variety of ways, and they feel it on a daily basis in everything from the meager level of attention they get from providers to the poor performance of the applications they use. They are not happy about it one bit.
One thing that has stood out recently in these conversations is how much displeasure there is about a very specific aspect of the technology infrastructure—their WAN.
Why the Dissatisfaction with WAN Technology?
Why the sudden dissatisfaction with a technology that has been a cornerstone of their network strategy forever? I think it comes to the evolution of our collective expectations for what a corporate network should be able to do. It didn’t happen overnight. I believe it happened in a steady way over the past several years and is now reaching a critical mass. One of the ways that expectations have changed is about uptime. In the days when “four 9s” was the general expectation, there wasn’t as big a spotlight on downtime. After all, four 9s allowed for up to 4 ½ minutes of no service a month. In that environment, the mercurial performance of many WANs was still within the realm of normal and acceptable, but expectations have evolved and now people see “five 9s” as the new standard. That equates to less than half a minute of downtime a month, which people naturally round down to zero. It’s only human to do so. Under that brighter spotlight, a WAN’s performance suddenly becomes a major pain point. Every customer call that comes in via a WAN and gets cut off by a technical glitch gets magnified. Every second that systems aren’t up starts to be seen as lost dollars. All of these issues existed before. What changed was expectations.
Expectations have also changed dramatically when it comes to issues like latency, installation times and cost. The inherent latency characteristics of a WAN that used to be A-OK now cause CIOs of mid-market companies to tear their hair out. Companies are also no longer willing to wait weeks or months for a loop to be installed to support branch offices. And the same holds true for cost. Costs that used to be acceptable now feel sky-high to companies, especially when you factor in static budgets and ever-growing bandwidth needs.
SD-WAN Offers a Win-Win Alternative.
That brings me back to the quest of a win-win. Companies that are fed up with their WAN do have an alternative that addresses these limitations, and it happens to be a solution that service providers can be enthusiastic about as well: SD-WAN. Software Define Networking (SDN) and Network Function Virtualization (NFV) allow companies to move away from traditional WAN technology and address so many of the complaints I mentioned above. For example, it allows companies to avoid costly branch office solutions and virtualize the customer edge (and beyond). SD-WAN also allows companies to get past the limitations of static routing tables and enter the world of sub-second, dynamic routing of packets based on policy. Last but not least, it also allows companies to lower costs in some significant ways. Simply put: customers get full visibility of their entire network via a single pane of glass, they define customer routing policies based on the way they run their business, and better application performance results at a lower TCO.
The benefits are also significant for service providers. The operator who utilizes SD-WAN technology gets to realize capital and operating cost efficiencies through “whitebox” networking, virtualization, and service automation. Service providers also benefit from the increased agility, reusable cloud infrastructure, and cost savings of SDN/NFV.
Seems like a win-win. But how fast will customer adoption follow? And what types of challenges will such implementations bring? I’m definitely interested to watch this unfold as customers and service providers continue to grapple with these challenges and find a new way forward