Improve your hybrid workplace video experience: Proven approaches for multi-cloud application performance management

Unified communications adoption is growing for obvious reasons. By the end of next year, it’s expected that 70% of employees will rely on an all-in-one collaboration app like Webex or Microsoft Teams to remain effectively connected and productive. This increasing reliance on cloud communications makes real-time performance management vital especially for video, a company’s “face forward” for digital business services and online customer engagement.

Now more than ever, companies need guaranteed performance between their enterprise network and the UC provider’s service network, which is critical to the delivery of optimal voice and video conferencing experiences from everywhere. To do this, they must focus their attention on performance management capabilities.

Before we explore these capabilities, let’s first consider the challenges that are affecting UCaaS application performance. Bear in mind that delivering an excellent (dare we say “perfect”?) video experience is much more difficult in today’s multi-cloud IT environments, especially with more employees bringing their own connections to the table.

UC application performance problems: 2 common root causes

1. Homeworkers on unreliable Internet services and/or Wi-Fi networks: A recent study from Altman Solon found that 64% of IT decision-makers anticipate a growing reliance on public network access methodologies due to the explosion of remote workforces. Half are using only broadband network connections to the public Internet and agree that their application performance is insufficient as a direct result.  Almost all of those surveyed admitted that the cost savings of broadband didn’t justify the degraded quality of service.
2. Routing remote users over enterprise VPNs:  Industry analysts at Metrigy do a good job of breaking this down in their new “Work-from-Anywhere Done Right” guide.  Many IT leaders assumed the shift to work from home would be temporary, so they relied initially on VPNs to provide secure remote worker access to data and application services. Unfortunately, VPNs (in conjunction with proxy servers and security devices) are significantly challenged to both scale and support the performance needs of increasingly critical collaboration technologies like video conferencing.

In fact, remote worker VPN connections often create inefficient network routing patterns (commonly referred to as “hairpinning”) that ultimately degrade the UC user experience. Here’s how it happens. VPNs add additional latency to the already present risk of jitter and packet loss by routing session traffic that is ultimately Internet-bound over a VPN connection to the company WAN, then back out Internet drain points to reach services platforms. Hairpinning techniques are problematic when compared to letting individual users access business-critical cloud applications directly via their own Internet connections (split-tunneling).

Delivering highly reliable video services: Performance management is key

Here are several ways organizations can address these challenges to achieve optimal UC performance.

Evaluate your network and your UC provider’s SLAs using the below criteria

• Uptime: Uptime is the amount of time (in days, hours, and minutes) the server, network, or website has been running (a.k.a. up). This is typically expressed as a percentage of the measurement period.
• Mean Opinion Score (MOS): MOS is a standard metric used to measure perceived voice call quality. Scores range from 1 to 5, with 1 being the lowest score and 5 the highest for excellent quality.
• Jitter: Jitter represents variations of delay in packet delivery, measured in milliseconds. The lower the jitter metric, the better; sub-millisecond jitter is outstanding.
• Latency: Latency is the delay between network entry and exit points, measured in milliseconds.  Latency outcomes are influenced by the geographical path of the route, network congestion/contention, and systems that are in the path of the packet.
• Packet loss: Packet loss is typically measured as a percentage of all packets transiting the network fabric.  In the ideal network, real-time traffic (Voice & Video) should not suffer any packet loss. Unfortunately, that is not a realistic expectation of the public Internet, where the provider platforms lack QoS enablement and are often over-subscribed. Still, absent of the challenges caused by over-subscription, it is not unreasonable to expect a performance level under 1%.
• Global footprint: This is the network topology, inclusive of each network point-of-presence (PoP) where traffic enters and exits a given provider network, along with the global reach of the unified communication service.

Manage the public internet: It’s possible and essential

The unpredictability of the public Internet, resulting in additional delay and/or packet loss can be frustrating to the user community when trying to use collaboration applications like Webex, or Microsoft Teams calling plans.  With a growing number of employees working from home, it’s vital to provide the best possible user experience for maximum participation and collaboration in a hybrid work environment. A properly configured SD-WAN routing engine continuously monitors the link for latency, packet loss, and jitter, using those metrics to steer application traffic across the ideal transport medium, as determined by the client’s requirements for performance.  This level of application-aware routing sophistication, coupled with quality transport services and path conditioning ensures superior performance for real-time apps like video conferencing.

Application prioritization, traffic steering, and path conditioning capabilities result in a highly effective and efficient network construct, optimizing the network spend while protecting the user experience.  With SD-WAN, direct connections to cloud service platforms from providers like Amazon Web Services (AWS) and Microsoft Azure can be established in multiple forms, as indicated by business criticality and the fault tolerance of the user. These attributes help simplify cloud migration strategies and access resiliency, dramatically improving application performance (more on this here). A fully managed service takes this a step further, providing performance monitoring services and security add-ons, including perimeter firewalls, continuous threat monitoring, and incident response services provided by a team of certified security analysts.

Here’s another SD-WAN tip: With so many companies relying on SD-WAN routing with Internet circuits in their WAN infrastructure, a new innovation makes these affordable connections perform like the most expensive ones do — delivering services more like the highly reliable, private Ethernet connections. Ask your network services provider about performance-boosting technologies they may offer to minimize packet loss and deliver enriched experiences across broadband connections.

Manage cloud performance: Network service providers can help you

As your organization relies more on third-party applications hosted in the cloud, your on-ramp to the cloud becomes more important than ever. Cloud SLAs add an additional assurance to boost confidence in UCaaS and cloud app performance. To be clear, cloud SLAs are not an agreement between a client and their cloud service provider (like Microsoft Azure or AWS). It’s an agreement between the client and their network service provider, who ultimately carries the traffic and hands it off to Microsoft Azure or AWS.

Cloud SLAs help organizations to select providers that can ensure consistent, high-quality service performance outcomes, especially when the application user community is dispersed globally.  Still, many IT leaders are led to believe that SLAs don’t matter in the cloud, or don’t apply to direct cloud connections. Further, most providers don’t even offer cloud SLAs, but the fact is that every service is an opportunity for an SLA. Masergy’s cloud SLA, for example, includes guarantees of 99.99% availability with redundant cloud connections and infrastructure, ensuring a more stable and resilient service experience.

Transform your VoIP and video experience

These approaches will help you transform your cloud UC application experience, enabling you to improve uptime, MOS, and SLAs for outstanding performance.  A huge part of this comes down to your network services provider.  As we’ve covered, your provider should offer uptime SLAs, cloud SLAs, reliable global availability, 100% packet delivery, and 24×7 monitoring and performance visibility.

Masergy delivers all of this and more. Learn more about our UCaaS offering