At ServiceNow, our work makes the world work. Our customers have high expectations, and we never want them to experience performance issues or system outages. Our customers’ success is our success, which is why it’s imperative their Now Platform instances run 24/7 with optimal performance.
Our customer escalations team is responsible to “run to the fire,” says Ash U., director of account escalation engineering. Historically, the team was called in to put out fires. Working reactively, they manually determined the most pressing issues for their engineers to prioritize to get the situations under control.
In regular analysis of customer escalations, the team noticed performance issues were increasing. This was due in part to the expanding number and size of customer instances and more complex customizations, some of which didn’t align with best practices.
Wanting to be proactive, not reactive, in responding to escalating issues, the team needed to identify trends and root causes—“smoke”—in order to prevent fires.
Two years ago, Ash and his team were tasked with detecting some of these “house on fire” situations before they spread. “We worked with three other teams: monitoring, SWAT [software action team], and support,” Ash says. Together, the teams analyzed how they were dealing with customer outages and what they could do to get ahead of them.
At first, the collaborating teams debated on what the solution should look like and whether it was possible to do what they were trying to achieve. Brian L., senior vice president of global customer support at ServiceNow, told the teams to pivot their thinking from "reasons it couldn’t work” to “what might work.”
That was the turning point. Emboldened with this fresh perspective, the teams developed their first predictive model: Predictive Performance Alerting (PPA). Based on monitoring alert data with a heightened level of sensitivity around certain alert types, the relatively simple model provided an effective early warning system for customers showing signs of smoke.
Since that time, Ash and his team developed two additional predictive models on the Now Platform, one of which uses ServiceNow machine learning and AI capabilities. The models use data to identify customer instances at risk of escalation. This allows Ash and his team to proactively take action before a situation deteriorates.
One model, trained using a historical set of escalation data records, looks at monitoring alerts. It provides nominations the other two models don't pick up on. “The three models are doing amazing things for our customers,” Ash says.
“A common talk track involves us saying to the customer, ‘We've got your back here, and we're ready to work with you to get your instance to a healthier state.’ We get them healthy again and, in doing so, we've prevented hundreds of outages so far,” Ash says proudly.
Any outage is inconvenient and can negatively affect customer service, employee productivity, and the business’s bottom line. “We’ll always be here to run to the fire, but the world works better when we can prevent fires,” Ash says.
ServiceNow CEO Bill McDermott often says, “Trust is earned in drops but lost in buckets.” The escalations team continues to engage early with customers to help ensure optimal instance performance. In doing so, the team builds trust and confidence, helping customers realize the true value of the Now Platform.
Join a proactive company that puts customer success first to make the world work better for everyone. Explore ServiceNow careers.
© 2023 ServiceNow, Inc. All rights reserved. ServiceNow, the ServiceNow logo, Now, and other ServiceNow marks are trademarks and/or registered trademarks of ServiceNow, Inc. in the United States and/or other countries. Other company names, product names, and logos may be trademarks of the respective companies with which they are associated.