The Automation Gap: Why ServiceNow Finally Did to Itself What It Sold to Everyone Else

Enterprise software companies have spent three years telling their customers to automate workers out of jobs. Most of those companies did not automate their own.

ServiceNow just published the most detailed account of what happened when it applied the same logic to itself. The numbers are specific enough to be worth reading twice.

The company grew from roughly 14,000 employees to nearly 30,000 over the past several years. Its operational support organization — the IT service desk, the HR operations team, the finance commissioning process — did not scale proportionally. The mechanism was not headcount reduction. It was what Jacqui Canney, ServiceNow's chief people and AI enablement officer, describes as capacity reallocation: the same people, doing fundamentally different work, at roughly 2.5 times the output.

HR business partners went from serving approximately 400 employees each to 1,000. The commissioning process for sales quotes — the back-and-forth between a seller and the finance team that once took an average of four days — now resolves in eight seconds. IT service desk tickets move from first touch to resolution autonomously in roughly nine out of ten cases. Of the staff who previously handled that work, 85 percent moved into SecOps, AI Ops, and Executive Briefing Centers; the remaining 15 percent now manage the agentic workforce itself: monitoring, intervening on edge cases, governing the system rather than triaging individual tickets.

"We grew from 14,000 to nearly 30,000 employees without a proportional increase in operational headcount," Canney said in a podcast interview at ServiceNow's Knowledge conference in Las Vegas. "It did more than double the output of what our people could do in people operations to serve the company as we were growing."

The governance layer that makes this coherent — AI Control Tower, which monitors every agent running across the organization, tracking adoption, token cost, performance, and security — emerged from an internal problem, not a product strategy. As agentic tools proliferated across ServiceNow's own teams, ungoverned agents created token cost spirals, duplication, and security exposure. Chief digital information officer Kellie Romack's internal answer became a product because enterprise buyers said they had the same problem. ServiceNow's AI Control Tower is now part of a broader enterprise security push: a global energy company cut threat containment time by 97 percent using the Autonomous Security and Risk product, and a major US financial institution eliminated 96 percent of its dormant non-human identities — the kind of unmanaged service accounts that become attack vectors when agents start proliferating across an organization.

Customers spending more than $1 million annually on ServiceNow's AI layer grew more than 130 percent year-over-year in Q1 2026, according to the company's financial results. Q1 total revenue was $3.77 billion, up 22 percent year over year. Bill McDermott, ServiceNow's chairman and CEO, told analysts the AI growth was "far exceeding even our own expectations."

The implications for the broader enterprise software market are concrete. ITSM and HR SaaS vendors whose platforms do not have a comparable redeployment architecture face a specific risk: as ServiceNow demonstrates that internal agentic transformation is operationally and commercially viable, enterprise buyers who have not retooled their own workforces look like upgrade candidates. Mid-tier vendors without a governance layer that can show similar workforce transition data face a platform eviction conversation.

The skeptical view is direct. Every number in this account comes from ServiceNow. No independent auditor has verified the 90 percent autonomous resolution rate, the 85 percent redeployment figure, or the eight-second commissioning benchmark. The governance narrative is also a product announcement. Redeployment only works if higher-value roles are actually available in the labor market — a condition that held for ServiceNow because it was simultaneously growing headcount by roughly 16,000 people. A company that is not in that growth phase, or whose affected employees lack the skills required for SecOps or AI Ops, faces a structurally different math.

The structural point survives those caveats. The gap between selling automation and living with it is real, and ServiceNow is the first major enterprise software company to close it publicly and with enough specificity to evaluate. The follow-on question — whether competitors like Salesforce, SAP, or Oracle are running comparable internal agentic deployments — is one those companies have not answered publicly. The silence is itself data.

What ServiceNow has published is an internal audit, run on its own systems, with its own employees, over a period when the company was also growing at a pace that makes the before-and-after comparison meaningful. That is a different kind of evidence than a benchmark claim or a product demo. Whether the numbers hold when other enterprises attempt the same transition is the next question. The answer will define the enterprise AI market for the next several years.