The key to weathering a major spike in volume like Super Bowl Sunday is not to scale your headcount but rather to create an environment in which agents never see the same repetitive "is it down" question. We utilize our CRM as a proactive notification engine instead of a passive database; by the time a user is even thinking about contacting us, they should already have received either a direct notification or a banner in their interface indicating that we are aware of any issues.
One tactic that has proven to be successful over time is the integration of Atlassian's Statuspage into our live chat widget within the CRM. We have set a rule wherein any 'Degraded Performance' or 'Major Outage' status listed on the Statuspage automatically generates a proactive alert to every user that opens the chat interface. This is effective because it reaches users at the moment they intend to ask a question. Rather than waiting in a queue to have their simple question answered, they immediately receive an automated message informing them that our engineering team has already begun to address the problem.
Through the use of audience filters, we target our proactive alerts specifically to 'Active Session' users in the area affected by an issue. By doing this, we prevent sending too many alerts to users who aren't experiencing issues, thereby maintaining the urgency of our messaging. The combination of real-time status syncs and targeted filters allows us to deflect as much as 40% of the volume of incoming support traffic during a major outage (enabling support to keep the lines open for the more urgent and complex inquiries that need to be handled by human agents).
