Design Better · The Curiosity Department, sponsored by Wix Studio

Bonus Episode: Amy Lokey, Chief Experience Officer at ServiceNow, live in SF for UserTesting's THiS Connect Tour

·32 min·2 clips

Amy Lokey reveals how ServiceNow's AI agents orchestrate teams like a manager to handle complex enterprise tasks autonomously.

1. Design Better host Eli Woolery interviews Amy Lokey, Chief Experience Officer at ServiceNow, in a live session recorded at UserTesting's THiS Connect City Tour in San Francisco on May 1st. 2. Lokey's career path runs from consumer products at Yahoo through LinkedIn (professional social network) and Google Workspace (then G Suite, where she led UX for productivity tools) to ServiceNow, where she has been for about five years. 3. At ServiceNow, her UX organization has grown to nearly 1,000 people, distributed across the United States, India, Europe, Egypt, and recently Costa Rica. 4. The episode's core thesis is that enterprise AI design requires managing trust, transparency, and multiple simultaneous stakeholder needs — challenges that differ fundamentally from consumer product design. 5. Lokey traces ServiceNow's AI evolution: generative AI skills (summarize a case, generate an email, write code) came first, then single-skill agentic automation (removing the human trigger), then an AI orchestrator that assembles teams of agents to handle complex multi-step workflows. 6. She describes the orchestrator as a 'team manager': when a trigger fires, it assembles the right combination of AI agent skills to complete a complex task, with humans approving write-actions while AI handles read-and-analyze work autonomously. 7. The next phase, agent-to-agent frameworks, will let ServiceNow's AI agents collaborate with agents from other platforms — she mentions a Microsoft Copilot integration announced the prior year, where a Copilot agent in Microsoft Teams could call in a ServiceNow service agent for IT support requests. 8. On enterprise AI design challenges, Lokey emphasizes that hallucination is far more serious at enterprise scale, driving investment in testing with realistic customer data and a staged 'innovation lab' early-access program before broad release. 9. Her team's branded UX quality metric combines time-on-task, accuracy, number of clicks, and qualitative ratings into a benchmark applied to every major ServiceNow product release twice a year. 10. She gives the field service management product as a case study: six or seven benchmarking rounds lifted usability from roughly 40% to about 85% — consumer-grade — with product revenue following the same upward trajectory. 11. ServiceNow uses internal tooling to log every research insight as an individual record, link it to a design artifact, connect that to an engineering epic and story, and track whether it shipped — creating end-to-end insight-to-impact accountability. 12. UX health scorecards, including insight burndown rates, are reviewed by executives in quarterly product reviews, giving UX teams direct visibility with leadership. 13. Lokey describes visiting SCAD and asking students to show how they use AI: a fashion design student demonstrated building a physical pink dress and using Midjourney to iterate dozens of variations — changing sleeves, length, silhouette — in hours rather than weeks. 14. She frames AI as an expansion of creative tooling analogous to past shifts: from Illustrator to Figma for UI design, AI is a 'same kind of tool but exponential,' reducing time spent on craft execution while still requiring the human idea and vision. 15. On the future of the design role, Lokey says 'we all move into more of the role of a creative director and visionary' — requiring the ability to articulate ideas to both AI models and human collaborators, and to evaluate whether outputs meet human needs. 16. For enterprise stakeholders, Lokey uses a three-altitude framework: C-suite (strategic business outcome), service owner/buyer (ROI and adoption), and end user (daily usability) — all three must be satisfied simultaneously by the same product. 17. She grounds the three-altitude framework in specific customer work: CVS focusing on reducing retail store turnover and Stellantis managing car supply chain from manufacturing to dealer delivery. 18. The episode is a live stage interview with an audience present; the format is conversational and includes one audience question about which types of insights resonate most with enterprise stakeholders. 19. Enterprise UX practitioners, design leaders, and UX researchers who work in or aspire to work in large-scale enterprise software will find this episode directly applicable. 20. Listeners focused on consumer products or early-stage startups may find the enterprise governance, scale, and stakeholder management content less immediately relevant.

As heard by us

A practical enterprise AI discussion about agents, Microsoft integration, testing, and the harder design burden inside business systems.

Design Better uses Amy Lokey's ServiceNow work to get past the usual enterprise AI demo talk and into how these systems might actually operate inside large companies.

Read the full review in PlayNext →

Why you'd press play

Listen for a measured look at how enterprise AI design has to earn trust in real systems.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Design Better