Now that every kind of user has had a dose of lockdown-mandated video conferencing – and new legions of homeworkers are suddenly reliant on United Communication and Collaboration (UCC) technology for meetings, collaboration and customer engagement – the honeymoon period is well and truly over for putting up with dodgy call quality.
Providers with proactive call analytics in place stand to gain the most as demand increases for services that ‘work no matter what’, irrespective of location, network, IP endpoint and time of day.
Everyone wants optimum call quality, but knowing what’s necessary to make it happen can be a frustrating, time wasting guessing game – especially now with home-based infrastructure sitting outside the control of corporate IT departments. And besides, we’re all armchair experts on call quality diagnostics now! Lately, who hasn’t pleaded with their family members to stay off Netflix or avoid hammering the WiFi during your crunch video calls? Or flatly blamed them for your frozen screen without knowing for sure it wasn’t caused by something else?
Asking for opinions misses the point
The domain of call analytics is undoubtedly a dry, back-office field of technology, but also one that everyone seems to have a strong opinion about – particularly when experiences are subpar. For instance, users of many video conferencing solutions will be familiar with the 5-star ratings engine that polls your experience following every session; the intention being to correlate scoring patterns with known technical issues. The invisible contract with the user is clear: tell us when it’s bad and this will go some small way to making it better next time.
In corporate environments, such as a contact centre implementation with hundreds or thousands of users, frequently polling users gets in the way of their productivity. Such reactivity – rather than proactivity – may also betray a failure of preventative maintenance and continuous improvement. In other words, what does it say about quality assurance standards within the UCC system if affected users must be relied upon to tell you that the experience isn’t good enough?
Fix the issue before the user notices
Gaining access to real-time and historical call metrics is imperative to recognising patterns and remediating issues before users become aware. But the key point here is ensuring that enough data is being collected across the environment to isolate the full range of potential causes. It may be the application itself is buggy, or that the underlying network is at fault. And finding all this out after the fact is small consolidation and doesn’t help IT teams keep to the their Service Level Agreements (SLA).
Looking at individual sessions in terms of standard based quality scoring (based on sampling, rather than reliant on user feedback), packet loss and jitter is a fundamental requirement. Then the wider enterprise network from the point of view of server, gateway and SBC performance metrics, as well as the endpoint client.
At the UCC softphone client, there are a host of additional metrics to check against, such as what kind of network is being used to connect the endpoint device e.g. WiFi, cellular, LAN. Also, is the softphone connected to headsets, external mics or speakers, and what specific type of hardware is being used?
