In the early days of the COVID-19 lockdowns, all contact centers suddenly became remote contact centers. Around the world, contact center managers scrambled to find tech solutions to help workers keep answering calls from home. In the rush to WFH, the experience for agents and callers was not good. Distorted voices, gaps, and dropped calls are all too common. These problems are usually not the fault of the contact center solution, but rather the inherent limitations of the internet.
How We Got Here: The Evolving Internet
The internet is built for the reliable delivery of content, such as email and file transfers, but not built for performance. Border Gateway Protocol (BGP)—the standardized routing protocol for the internet—has no idea what traffic it is routing or in which direction it is sending data. It simply routes data using the cheapest path possible depending on various factors, including politics, commercial agreements, and cost. As a result, certain types of data, namely real-time communications, suffer as the internet favors servicing volumetric traffic.
At the start of the millennium, content delivery networks employed global networks of servers to deliver static content faster, leading the way to make OTT (Over-the-top, services offered over the top of traditional content distributors) possible with the infrastructure built by companies such as Netflix, YouTube, and Akamai. More recently, companies like Amazon Web Services have democratized cloud computing, bringing connectivity to more users in more regions of the world.
The next evolution of the internet should be one in which the network itself is democratized, enabling speed-of-light communication for the applications that require high-velocity delivery, such as real-time communications. This evolution means delivering voice over internet protocol calls (VoIP), video and more, regardless of the compute or data server owner.
Complex Issues with a Complex Network
The internet is a complicated web of networks and billions of computers and devices that connect and work together to send and receive information. The greater the distance between the sending and receiving computers, the longer it takes the data to arrive at its destination. Information can also take different paths, which may increase or decrease the speed of delivery.
A common analogy used to explain how the internet works is the postal service. In the same way that your letter travels from you to the recipient through multiple facilities (carrier, sorting, post office), data travels through numerous routers on its way to the recipient computer. Just as your letter can arrive faster depending on which mail service you use, so too can data transmit faster depending on which path it takes, though the distance between the sender and receiver has the most influence on travel time.
There are, however, a couple of significant differences in this analogy. There are generally more “stops” along the way for data. Also, the data is broken up into multiple smaller “packets” of information rather than sent as one whole mail piece.
Because there are different paths along which individual packets of information can travel, each entailing a variable number of stops along the way, packets of data can easily get lost.
Packet loss, as it’s called, results in missing information. In a VoIP call, users experience packet loss as gaps in audio. Packet loss results in strange behavior like freezing or pixelated video in video chats.
Another result of these inconsistent, differing routes is variation in delivery time. Some routes are faster than others, so some packets may arrive more quickly since the internet is not conscious of which path it takes to send data from one point to another. This variation in the time it takes for data to travel across the network from one endpoint to another is called jitter. Jitter manifests as jumbled speech that creates an indecipherable conversation on a call.
While lag (high levels of latency), jitter, and packet loss may impact any online service or application, specific applications depend upon real-time packet delivery and minimal jitter and packet loss to function correctly and maintain a reasonable quality of experience. Packet delivery, consistent latency, and minimal lag are critical for services and applications like VoIP and video chat and in numerous other industries such as online investment platforms, telemedicine, remote collaboration, virtual desktops and remote contact centers.
For these applications, milliseconds saved in delivering data can lead to a real or virtual life-and-death situation and mean the difference between quickly addressing a customer’s issue and a frustrating interaction.
Why the Internet Doesn’t Work for Remote Contact Centers
There are four primary reasons why the internet doesn’t work for your contact center and why it will never be the answer for today’s real-time applications:
- The cables that comprise the internet are not wired for your communications. Because a large part of the original internet’s content comes from U.S.-based entities, you can see that many undersea cables are directed to connect countries to the U.S. rather than countries to each other in a mesh pattern. As the map below shows, if your agents are in India or the Philippines, it’s likely that your Australian customers would route through the U.S. Direct paths are more in demand today but come at often unreasonably higher costs.
Cable map. Source: http://www.submarinecablemap.com/
- The internet is an amalgamation of different service providers (xSPs) and Autonomous System Number/Networks (ASNs) joined together through Border Gateway Protocol (BGP) routing policies. While it is in the interest of an xSP to minimize the number of hops and cables that traffic must flow through, when it comes to ASN-to-ASN-to-ASN traffic, xSPs use BGP to facilitate traffic delivery, prioritizing the lowest cost possible through various peering and transit arrangements. That’s why your calls may not use the cable with the shortest path you see on the map above. On a less macro level, sub-optimal routing also happens inside a country and even cities. For example, it’s common to see traffic from one provider in the Philippines connect to another provider through a peering link in the United States. Though certainly not isolated, the diagram below shows a real-world example. The route from Atlanta to Washington D.C. was direct, but the return path went through California.
-
The BGP protocol isn’t responsive to outages and congestion the way you think it should be. Unless there is a broken link, BGP doesn’t care. That means your call could see a lot of data loss and jitter, and BGP doesn’t care because the vast majority of internet traffic isn’t sensitive to either.
-
Builders of infrastructure look to where the capacity needs are—in streaming video and CDN. Service providers focus on where to get the cheapest capacity and, in many cases, actively throttle network quality. Their “best effort” connectivity guarantee might just as easily be called “least effort.” As long as it works, as long as users get their files or watch a video with a few seconds of pre-buffering, it’s good enough.
What “Good Enough” Means for Remote Contact Centers
The result of all those factors that make the public internet ineffective? Lag. Lag is a general term used to describe long and/or unstable latencies, i.e., jitter, and packet loss, resulting in inconsistent application behavior and poor call quality.
The latencies between your agents and your customers may be permanently slower than what is ideal and necessary. It may be so for periods that last from a few hours to several days, or calls may experience brief periods of congestion, loss, and jitter that negatively impact call quality.
A customer experiencing lag does not enjoy the call. When calls are distorted, cut out, or drop altogether, this not only negatively affects the quality of the experience for the caller and agent, but also affects the perceived value your company places on its customers.
One of the first places a contact center is likely to see problems is in a low Mean Opinion Score (MOS). MOS, the most common metric used to measure voice and video call quality, ranges from 1 (bad) to 5 (excellent). To improve your MOS score, it is important to provide your customers with not only exceptional QoS, but also QoE.
According to Qualinet White Paper on Definitions of Quality of Experience (2012), a research project by the European Network on Quality of Experience in Multimedia Systems and Services, “Quality of Experience (QoE) is the degree of delight or annoyance of the user of an application or service. It results from the fulfillment of his or her expectations with respect to the utility and/or enjoyment of the application or service in the light of the user’s personality and current state.”
The bottom line is that the lag, latency and jitter on the public internet leads to frustrations by both the consumers and the call agents due to dropped calls, poor call quality, bad video reception, and lag. To solve this problem, the world needs a better, more simple solution.
How Subspace Improves Real-Time Application Delivery
A millisecond’s delay in delivery does not matter in sending an email or a text message, but it does matter in real-time communications.
Over the past three years, Subspace has deployed custom hardware in hundreds of cities and uses its proprietary AI software to “weather map” the internet. This infrastructure gives Subspace the power to find the best paths, a combination of existing paths and Subspace’s proprietary fiber-optic backbone, and pull traffic through those paths. It is like a private carpool lane and GPS, but for dynamic internet traffic. As a result, latency and lag are reduced, and real-time applications reach their full potential.
While other companies have attempted to provide solutions for expediting real-time communications on the internet, none have addressed the issue across the entire system. Many introduce other problems in their attempts. Commercial networking solutions like Cisco and Juniper were not built for global coordination or to understand and control communications traffic. Standalone solutions that require SDKs installed in clients and servers can create significant security and stability risks. And services provided by CDNs like Cloudflare and Akamai utilize the same principles and infrastructure as their primary businesses—that is, volumetric traffic—making the possibility of optimizing real-time traffic at scale a costly and difficult endeavor.
Subspace optimizes every component of network performance, from the infrastructure stack to the networking stack, including the control plane, and the data plane. With a global internet architected specifically for real-time traffic at every point, we are achieving speed-of-light communication and democratizing the network.
Regardless of the data’s location, Subspace uses a vast system of internet quality metrics and a proprietary algorithm to direct real-time interactive traffic onto our platform and across the public and our private internet to and from servers. The algorithm can balance multiple variables to suit each application’s needs—variables such as absolute latency, any latency above a threshold, jitter and loss. The optimal delivery platform via the Subspace onramp drives better, more consistent experiences that match each application’s needs.
Accessing Subspace
The world’s fastest, most stable, highly available and secure network is now publicly available. Subspace provides real-time application developers and companies with a quick way to operate, deploy, and scale their applications. We deliver real-time connectivity from anywhere to anywhere, and now, you can get enterprise-level connectivity anywhere in minutes, with no client-side installation required.
Until now, Subspace has worked exclusively with large companies on their particular network needs. But now, Subspace’s WebRTC-CDN brings the capability of ubiquitous private networking to every application, developer, and internet user—expanding the possibilities for truly instant digital experiences.
Subspace WebRTC-CDN enables real-time applications to perform everywhere in the world, without client-side appliances, VPNs, or custom software. Real-time voice and video applications, multiplayer online games, modern fintech solutions, transportation apps, and every internet application in between have access within minutes to the world’s first WebRTC delivery network. You can get your contact center traffic onto our highly available, ultra-low latency, global private network with a simple configuration change.
Easy Access to Subspace: The World’s First WebRTC Delivery Network
Subspace provides companies the capability of ubiquitous private networking for every real-time application. Developers can build instant digital experiences unencumbered by network constraints. Features of the dedicated network enterprise companies use are now accessible, self-service, by any internet application.
- Improved quality of experience. Faster connectivity and consistent network quality mean fewer dropped calls, more satisfied customers, happier agents, instant remote collaboration, and truly real-time experiences. On Subspace, you see a reduction in jitter, packet loss and latency by greater than 80%. With the Subspace SLA, we guarantee network accessibility and uptime. When your contact center operates on Subspace, you will always have access to the first network designed for real-time communication. When you implement WebRTC-CDN for your contact center, you can improve your MOS Score by up to 20%.
- Scalability and P2P Acceleration. Unconstrained by network limitations and with company hardware management removed, companies can scale their remote worker footprint, quickly react to increases and inconsistencies in demand, and create instant digital experiences. Subspace simplifies and accelerates your P2P connections, giving your voice and video users the low-latency connections they need with a simple API call.
- 100+ Global PoPs. With more than 100 global PoPs and growing, and high interconnection with global providers, Subspace puts your audience just milliseconds away from the global internet. No longer will it matter that your home office is in the Western U.S., your contact center is in India, and your caller is in Australia. Your audience is milliseconds away from the internet, enabling true real-time interactions.
- Anycast Optimization Eliminates Dropped Calls and Packet Loss. Leveraging Subspace’s proprietary AI-based anycast network analysis, Subspace routes users across the globe onto the entry point nearest their location. Continuous monitoring ensures traffic remains on the most optimal path, even when internet interference occurs. With dynamic packet route, packets travel to the destination in real-time with no packet loss. Our drop-free anycast routing means that every real-time interaction survives a network path change.
- Higher profitability. With improved end-user experience, customer churn decreases, more customers adopt the product, and revenue increases, while fewer requirements for company hardware and management decrease costs, altogether increasing profitability.
- Security. With Subspace, you get the peace of mind that your contact center is safe and secure. You get high quality, low packet loss, with always-on, inline DDoS protection, with no additional latency. With Subspace WebRTC-CDN, we simplify enterprise-level security, giving your company, your call agents, and your customers the peace of mind they deserve.
Modeled after Subspace data. Depicting packet loss from Bangalore and Miami.
Today, many of us use the internet for conferencing, and many of us experience lag during our calls—delays between callers that cause disruptions or the audio/video periodically clipping out—but accept the mediocre quality as if it's an inevitable fact of internet life. However, in customer service, millisecond delays in the delivery of data create a frustrating situation in which the interpersonal skills of the best agents cannot beat laggy internet connectivity.
What’s Next for the Internet and Remote Contact Centers?
No, the internet was not built to support real-time communications traffic. But Subspace was. We are a dedicated network built for real-time applications. WebRTC-CDN is a managed WebRTC service that lets you quickly route your traffic over the world’s most highly available ultra low latency, global private network. With a simple configuration change, you immediately increase speed, accessibility and QoE, leading to a higher MOS score. . Your users instantly see the difference when your contact center routes traffic on the only network purpose-built for real-time applications.
When your customer experience depends on real-time connectivity, you need Subspace WebRTC-CDN.
