How Microsoft Teams handled the rapid surge in usage

On March 16, the first Monday when most of Europe began working from home, Microsoft’s Teams suffered a rare outage. The service interruption lasted for almost two hours. This foreshadowed the things to come – a massive influx of users to the digital platforms for communication and collaboration at workplaces, and even personal usage.

Microsoft Teams, like many other existing collaboration applications, has seen an exponential growth in usage as the world is staying indoors due to the Covid-19 pandemic. As of April 2020, Microsoft Teams has 75 million daily active users of which 31 million were added in just over a month. The platform has seen a whopping 70% growth. The company has also well leveraged this opportunity to roll out a wide range of new features to help accelerate the adoption. 

Unlike Zoom, Microsoft Teams hasn’t found itself surrounded in any privacy controversy. The platform has also been running smoothly with no big outage reported since March 16. So, how is it like for a company such as Microsoft to handle such an unprecedented growth of users. 

We spoke to Samik Roy, Country Head, Modern Workplace, Microsoft India on how Teams’ team has worked in the last few months. Edited excerpts.

What are the key factors that helped Microsoft handle the spike in usage of Teams?

As the demand for Teams surged, we at Microsoft quickly took steps towards managing increased cloud infrastructure and network demand. We optimized and load-balanced the Teams architecture and quickly rolled these out worldwide (using Azure DevOps), without interrupting the customer experience. This work is durable such that we can manage Teams rapid growth moving forward without creating pressure on Azure customers’ capacity needs. We proactively expedited additional server capacity to the specific regions where we expected higher demand, while ensuring the safety and health of our datacenter staff and supply chain partners. We removed restrictions for new free and benefit subscriptions in several regions, so that anyone can learn more about Azure’s capabilities and develop new skills. In addition to that, we are refining our Azure demand models. Our data science models are using what we’ve learned from this pandemic to better forecast future demands, including adding more support to handle future global events like a pandemic that drives simultaneous demand usage everywhere in the world. We remain committed to operational excellence and we will continue to share what we are learning and doing to support everyone during this time.

Microsoft may have also been working with limited resources. How is the company managing the unprecedented demand and service complaints or requests?

We strive to limit the scope of all incidents by having multiple instances of each service partitioned off from each other. In addition, we’re continuously driving improvements in monitoring through automation, enabling faster incident detection and response. Just as the services are designed and operated in an active/active fashion and are partitioned off from each other to prevent a failure in one from affecting another, the code base of the service is developed using similar partitioning principles called fault isolation. Fault isolation measures are incremental protections made within the code base itself. These measures help prevent an issue in one area from cascading into other areas of operation.

Do you think digital tools are more relevant now than ever before? Do you think it's a short-term growth?

Every crisis has its learnings. COVID–19 triggered social distancing and lockdown forced organizations to look at remote working for business, learning and healthcare continuity. This has accelerated India and organizations’ pace towards true digital transformation. We hope that everyone can see how people can work remotely on a secured platform and collaborate effectively. Zero commute times have translated into higher efficiency and productivity for those who are in the remote working eco system. 

It was very encouraging to hear TCS leadership during the recent result announcement. They mentioned that their future operating model will have 25% of their workforce at their facilities and they will continue to drive high productivity with anytime, anywhere, collaborative working. During COVID-19, they adopted collaborative and cloud-based technologies like Teams to enhance business continuity and enable secure connections to their systems and networks.