ServiceNow reported third-quarter 2026 revenue of $4.106 billion, up 20.5% year on year, including $3.988 billion of subscription revenue (up 21%). The contract numbers looked solid too: current remaining performance obligations of $13.65 billion, up 20.5%, and total RPO of $29.7 billion, up 22%. The company raised its full-year subscription revenue and cash-flow outlook, and now guides to a 36% non-GAAP free-cash-flow margin.
The stock fell more than a quarter anyway — from about $150 the day before to roughly $110. The trigger was not the quarter but the look past it: ServiceNow guided fourth-quarter cRPO growth of 18.5%, where analysts had modelled about 19.8%. That extends a visible slide — 22.5%, 21% and 20.5% through the first three quarters of 2026, now 18.5%. cRPO measures revenue expected from contracts already signed over the next twelve months, which makes it the best leading indicator a subscription business has.
A narrow earnings miss added to it: $1.00 adjusted per share against roughly $1.02 expected. Part of that is bookkeeping — the Armis cybersecurity acquisition cost the company interest income because less cash was earning, and the borrowing behind it raised interest expense. That explains a cent, not a collapse.
The question the market is actually answering is what a company is worth when its growth decelerates at the very moment its own technology undermines the industry's billing model. Since the selloff nicknamed the "SaaSpocalypse" in July — set off by the fear that autonomous agents would reduce headcount and therefore seats — software valuations have been priced against that question.
ServiceNow's answer is notably double-edged. It raised its 2026 AI annual contract value target from $1 billion at the start of the year to $1.7 billion, and said production consumption of agentic AI is up more than thirteenfold since January. At the same time the classic subscription business is slowing. New contracts are arriving; they are not yet replacing what the old ones contributed to growth, or they are shaped differently — consumption-based rather than per-seat.
That is not a crisis, but it is a repricing. Anyone buying ServiceNow is betting that agents embed the platform deeper into operations and raise revenue per customer instead of cannibalising licences. The next two quarters decide whether 18.5% was a quirk of contract timing or the start of a slower normal. For the software industry as a whole, that number matters more than the share price.
