ServiceNow Raised Its AI Target to $1.5 Billion. The Market Sold the Stock Anyway.

ServiceNow beat its own numbers Wednesday. The stock fell 13 percent anyway.

The disconnect is the story. Chief Executive Bill McDermott spent much of the earnings call insisting that direct AI model approaches are parlor tricks that cost enterprise customers ten times more than ServiceNow's embedded AI (Business Insider). He said AI has never been a bigger tailwind for the company in its history. And yet investors sent the stock lower after the numbers landed — even as the company raised its full-year subscription revenue outlook to $15.74 billion to $15.78 billion, up more than $200 million from its prior forecast (The Motley Fool earnings transcript), and now expects $1.5 billion in AI-specific commitments for 2026, a 50 percent jump from the $1 billion target it set just months ago (Business Insider).

The market was not persuaded, at least not entirely, by the confidence. Several factors drove the selloff. The company blamed 75 basis points of Q1 subscription revenue headwind on delayed closings of large on-premise deals in the Middle East, a geopolitical drag that CFO Gina Mastantuono said prompted incremental conservatism in the full-year outlook (The Motley Fool earnings transcript). Royal Bank of Canada analysts called the 2026 outlook mixed when Armis is excluded from the numbers (Business Insider) — suggesting the $7.75 billion cybersecurity acquisition the company closed this quarter is doing meaningful work to prop up the guidance, work that has not yet been proven in production (CNBC).

That is the harder question ServiceNow has not yet answered: what happens when you strip out the acquisition? The company argues it is too early to judge Armis integration. McDermott called the deal a control tower for enterprise AI governance. Veza, the identity security company ServiceNow also acquired, is supposed to do for AI agent permissions what Active Directory never quite managed for human access. The vision is coherent. The execution timeline is not.

The numbers themselves remain strong. Remaining performance obligations grew 23.5 percent to $27.7 billion (The Motley Fool earnings transcript). The company closed 16 deals above $5 million in net new annual contract value, five above $10 million. Renewal rates came in at 97 percent. Deal volume for three or more Now Assist AI products grew nearly 70 percent year over year. The $1 million-plus customer cohort expanded by more than 130 percent. These are not the metrics of a company in trouble.

But they are also not the metrics of a company whose stock deserves to trade where it did in January, before the broader software-sector correction and before the Iran-related deal delays became a known quantity. ServiceNow is down roughly 30 percent year to date as of Wednesday (CNBC). The AI control tower narrative is compelling. The market is waiting to see whether the plumbing works at scale before pricing it in.

The broader enterprise software industry is watching closely. Salesforce, SAP, and Microsoft have each made comparable moves toward bundling AI capabilities into workflow platforms over the past year (CIO Dive). ServiceNow's experience is a leading indicator for whether the enterprise AI infrastructure bet pays off — or whether the gap between booking numbers and durable revenue stays too wide to close quietly.

McDermott is right about one thing: the demand is real. The $1.5 billion AI commitment figure is up 50 percent in a single quarter. What remains unproven is whether ServiceNow can deliver on it before the next earnings call forces another verdict.