NEA Blog

IIX: Network Interconnectivity for the Enterprise Masses

In today’s anytime-anywhere business world, we expect instantaneous response whenever we access a file from Box, search for a new recruit on LinkedIn, or launch a video conference on Blue Jeans Network. And in today’s mobile first, real-time data, video-centric, everything-as-a-service world, the costs are high when a business fails to deliver.  A widely quoted calculation when it comes to B2C applications is that a page load slowdown of just one second could cost Amazon $1.6 billion in sales each year. [1]  The impact of slow performance is arguably just as great for B2B applications like LinkedIn and Box, where sluggish response times can lead to lost productivity, missed opportunities, higher costs, and lower revenue.  The consequences can be even graver for video services like Blue Jeans, where small delays can render their application not just slow but outright unusable.

Internet giants and very large enterprises with high traffic volumes have both the means and the method to circumvent the performance, predictability and security issues that are intensifying with continued growth of Internet traffic.  But what about everybody else? It’s a critical question nobody has been able to answer—until now. IIX (International Internet Exchange), a next-generation software and interconnection platform that today announced a $10.4M Series A round led by NEA, is bringing network interconnectivity to the enterprise masses.

It can’t happen soon enough. Despite the massive economic and technological impact of the Internet, the underlying architecture connecting it all has changed little in the last decade. The typical method of connectivity used to traverse packets across the networks that make up the global Internet is growing increasingly ineffective for delivering real-time services from the cloud.  Internet traffic is exploding as SaaS applications further penetrate the enterprise, workers collaborate more online (especially via video), and everyone accesses more content on-demand.  As traffic grows, network congestion and slow response times are escalating, resulting in billions of lost dollars and unhappy customers.  This congestion is especially detrimental in a world in which applications are increasingly assembled from third-party cloud services.  For example, LinkedIn may serve up profiles more slowly to an end user if the display advertisements on those pages it is receiving from Google are impeded by public web congestion between LinkedIn and Google’s datacenters.  On top of all this, the explosion of activity on the web has attracted the attention of hackers seeking to disrupt cloud services through the use of denial-of-service and other network-based attacks that take advantage of the Internet’s “openness.”

The world’s largest Internet companies have gotten around these issues by “peering”—they operate their own private networks and directly connect to the networks of their business partners or customers at centralized colocation facilities known as Internet Exchange Points (IXPs).  Internet giants like Google, Facebook, and Amazon can afford to build their own networks by buying up dark fiber, while the typical large enterprise pays a Telco to provision a network for them based on MPLS (Multiprotocol Label Switching).  That MPLS network can be used to connect datacenters together within an enterprise’s own network or connect datacenters to IXPs, where they can directly connect with the networks of others.

That approach makes sense for companies with massive traffic volumes (e.g., consumer Internet) or guaranteed security needs (e.g, financial services and government), but it’s certainly not for everyone. Operating and interconnecting private networks is both complex and costly, requiring a tremendous commitment of time and resources—from hiring network architects to buying and managing equipment to wrangling with Telcos.

How is IIX changing the game? By virtualizing the IXPs at the core of the Internet to enable a customer to use software to directly connect within their own distributed network or to the networks of others.  This avoids the commodity web and delivers increased performance, reliability, and security for applications.  The service requires no capex and overall opex is much lower than paying for MPLS lines.  In one case, IIX enabled a customer to directly reach 10 regions while saving them over $1M in capex and 40% in monthly opex relative to operating their own private network.

IIX is already used by numerous emerging technology companies (including NEA portfolio companies Box and Blue Jeans Network).  It is also even used by leading Internet giants (such as LinkedIn, GoDaddy, and TripAdvisor), despite their capacity to run their own networks, since IIX allows these companies to directly connect to new regions faster and cheaper.  IIX’s ecosystem is growing rapidly—over the long-term, we expect IIX to grow into a “LinkedIn” for enterprise network interconnectivity, where an enterprise can connect their network to the network of a trusted partner or customer with a simple click.

We’re especially excited about working with the team at IIX.  CEO & Founder Al Burgio has assembled a world-class team of executives and advisers with proven success founding and running many of the largest IXPs, including Chief Strategy Officer Bill Norton (aka “Dr. Peering”), Equinix co-founder and the author of the “Internet Peering Playbook”; and Morgan Snyder, VP of Sales, the first Equinix sales rep and key employee who helped that organization scale over nine years and brings two decades of sales experience with telecom and datacenters. IIX’s advisory board is composed of leading experts in Internet architecture, including the prolific Internet entrepreneur Jay Adelson, a founder of Equinix.  We look forward to working with the IIX team on their mission to bring fast, reliable, and secure interconnectivity to enterprises everywhere.

[1] http://www.fastcompany.com/1825005/how-one-second-could-cost-amazon-16-billion-sales