We put Now Assist to work and saw value quickly
Now Assist is ServiceNow’s generative AI experience, built into the Now Platform so that businesses can accelerate productivity across all their workflows.
Our GenAI strategy is to learn, deploy quickly, iterate, and drive trackable business value.
Our Now Assist goals
We applied Now Assist at ServiceNow to meet several business needs:
• Reduce manual work for agents with summaries and resolutions to help them start work fast
• Generate content automatically, including intelligent search results, work notes, and knowledge base articles
• Deliver engaging experiences with natural human language that deflect cases by empowering people to self-serve
Accelerating agent productivity
For our initial implementation, we rolled out incident summarization and code generation in a pilot, got feedback, made changes, and rolled out to a broader audience. As a natural extension of that, we enabled resolution notes to reduce the time it takes agents to create notes documenting how a case was resolved.
With Now Assist, notes were generated within seconds and the agents needed only to review and refine the notes, saving approximately 80% of the time formerly needed for each resolution note.
Two months later, we extended Now Assist to support our IT agents further with chat summarization.
By month four, we were using Now Assist to generate knowledge articles, saving our IT agents even more time.
Then we extended Now Assist beyond supporting IT agents—to customer, support, and HR agents. Extending these GenAI capabilities to HR agents required ensuring robust privacy and compliance.
Within only 120 days, we were seeing value using Now Assist that was on track to enhance productivity equivalent to 50 FTEs per year across customer and employee services on an annualized basis. This increased by over 40% to $14.4 by the following quarter.
Enhancing employee experience
In addition to agent productivity, we also wanted to use GenAI to enhance our employee experience, making it easier for our employees to get answers. The more often GenAI could help employees self-serve, the more cases we could deflect to avoid agent intervention. For our GenAI experience journey, we followed the same strategy that we enacted to improve agent efficiency: learn, deploy quickly, iterate, and drive trackable business value.
Because we already had a robust knowledge base, we were able to launch Now Assist in Search. This provided summaries for questions asked, making it easier and faster to find answers and information.
Within a couple of months, we deployed this solution in our employee portal, My ServiceNow, where employees go to submit tickets. This positioned GenAI to offer smart search results and suggestions that helped customers self-serve, rather than opening a ticket. The result was an upsurge in GenAI assisted self-service and a 14% boost to employee deflection rate.
Now Assist in Creator was also a relatively low effort to turn on. It did not have a lot of prerequisites other than compatibility in our instance and integrating it into our workflow. It has improved our developer productivity. We are saving hours that are the equivalent to a 10% productivity gain and a code generation acceptance rate of 48%.
Now Assist quickly generated tangible benefit
Within months of deployment, Now Assist was being used across the enterprise, increasing productivity by deflecting cases and automating manual work while also improving our employee self-serve experience.
Within 120 days of implementing Now Assist, we had $10M in annualized tangible benefit, representing $5M+ annualized cost takeout and an additional $4M+ in productivity. Within months, tangible benefit increased 40% to $14.4M.
Today, employees in every department at ServiceNow use Now Assist to supercharge their workflows, transform experiences, and increase productivity. We are continuing to expand our use cases, deploy quickly, iterate, and drive trackable business value.